JP3779075B2 - Control device for internal combustion engine - Google Patents

Description

【０１０２】
ここで、式（１）中の係数Ｇa1は、あらかじめ定めた所定値（一定値）である。
【０１０３】
そして、この推定吸入空気量gair/preを、ＦＩＲＥモードの動作中に次式（２）により制御サイクル毎に累積加算していくことで、吸入空気量の積算値qair/preを求める（以下、この積算値qair/preを推定積算吸入空気量qair/preという）。
【０１０４】
【数２】

【０１０５】
ここで、式（２）中のｋは制御サイクルの番数である（以下、同様）。
【０１０６】
尚、内燃機関１の実際の吸入空気量の積算値は、例えばエアーフローセンサ等により制御サイクル毎の吸入空気量を直接的に検出し、それを積算することで得るようにしてもよい。
【０１０７】
次に、触媒装置３に与える熱量の積算値の目標値に相当する吸入空気量の積算値の目標値（以下、目標積算吸入空気量qair/cmdという）は、触媒装置３の昇温・活性化を適正に行う上では、種々の形態で設定することが可能である。しかるに、その目標値は、ＦＩＲＥモードの動作中の内燃機関１の吸入空気量、ひいては内燃機関１の燃焼状態、エミッション状態等に影響を及ぼすこととなるので、該内燃機関１の運転の安定性を考慮する必要がある。
【０１０８】
そこで、本実施形態では、前述の如く理想的な条件下で、触媒装置３の昇温・活性化を適正に行い、また、安定した内燃機関１の運転を行い得るように定めた前記標準開度指令値θ0 に基づいて上記目標積算吸入空気量qair/cmdを設定する。
【０１０９】
すなわち、前記標準開度指令値θ0 は、バイパス開度の指令値に対して実際のバイパス開度や吸入空気量が一義的に定まる、大気圧Ｐａ及び大気温度Ｔａがそれぞれ一定の標準大気圧及び標準大気温度である、というような理想的な条件下で、触媒装置３の昇温・活性化を適正に行い、また、安定した内燃機関１の運転を行い得るように設定したものである。言い換えれば、標準開度指令値θ0 は、触媒装置３の昇温・活性化を適正に行い、また、安定した内燃機関１の運転を行うために燃焼室４に吸入されるべき最適な吸入空気量を規定するものである。
【０１１０】
従って、上記のような理想的な条件下で、バイパス開度を標準開度指令値θ0 に従って操作した場合における内燃機関１の燃焼室４の制御サイクル毎の吸入空気量を、制御サイクル毎の目標吸入空気量gair/cmdとして設定し、その目標吸入空気量gair/cmdの積算値を上記目標積算吸入空気量qair/cmdとして設定すればよい。
【０１１１】
この場合、上記目標吸入空気量gair/cmd及び目標積算吸入空気量qair/cmdは制御サイクル毎に次のように標準開度指令値θ0 から求めることができる。
【０１１２】
まず、実際のバイパス開度をθとしたとき、バイパス弁７を通過する単位時間（一定時間）当たりの空気量Ｇｉは、バイパス弁７の上流側の圧力である大気圧Ｐａと下流側の圧力である吸気圧Ｐｂとを用いて一般的に次式（３）により表される。
【０１１３】
【数３】

【０１１４】
ここで、式（３）中のＣｉは、空気密度に応じた係数で、該空気密度は大気温度Ｔａに応じたものとなる。また、式（３）中のθの項は、厳密には、バイパス弁７の箇所の有効開口面積であるが、ここでは、この有効開口面積の代わりにバイパス開度θを用いている。尚、有効開口面積に代わりにバイパス開度θを用いることによる影響の補正は、係数Ｃｉに含めるようにしてもよい。
【０１１５】
この場合、前記の理想的条件下で、バイパス開度を標準開度指令値θ0 に従って操作したとき、式（３）において、θ＝θ0 、Ｐａ＝標準大気圧（一定）となり、また、係数Ｃｉも基本的には標準大気温度に応じた一定値となる。また、ＦＩＲＥモードにおける内燃機関１の定常的な運転状態では、吸気圧Ｐｂの変化は比較的小さく、概ね一定値となる。さらに、ＦＩＲＥモードにおける内燃機関１の定常的な運転状態では、前記スロットル弁５は基本的には閉弁状態で、内燃機関１の燃焼室４の吸入空気量は、バイパス弁７を通過する空気量Ｇｉに等しいと見なせる。
【０１１６】
従って、理想的条件下で、バイパス開度を標準開度指令値θ0 に従って操作したときの燃焼室４の単位時間（一定時間）当たりの吸入空気量は、標準開度指令値θ0 に比例する。
【０１１７】
よって、理想的条件下で、バイパス開度を標準開度指令値θ0 に従って操作したときの燃焼室４の１制御サイクル（１ＴＤＣ）当たりの吸入空気量、すなわち、前記目標吸入空気量gair/cmdは、次式（４）により求めることができる。
【０１１８】
【数４】

【０１１９】
ここで、式（４）中で内燃機関１の回転数Ｎｅの逆数（１／Ｎｅ）の項が含まれるのは、本実施形態で１制御サイクル（１ＴＤＣ）の時間が回転数Ｎｅに反比例するからである。また、式（４）中のパラメータＧa2は、前述したことから明らかなように、標準大気圧や標準大気温度、ＦＩＲＥモードにおける内燃機関１の定常的な運転状態での標準的な吸気圧Ｐｂ等により定まる所定の定数である。
【０１２０】
そして、本実施形態では、この式（４）により求まる目標吸入空気量gair/cmdを、ＦＩＲＥモードの動作中に次式（５）により制御サイクル毎に累積加算していくことで、前記目標積算吸入空気量qair/cmdを求める。
【０１２１】
【数５】

【０１２２】
尚、このようにして求められる目標積算吸入空気量qair/cmdは、標準開度指令値θ0 により定まるものであるため、内燃機関１の始動時の機関温度Ｔｗ（あるいは触媒装置３の初期温度）に応じたものとなる。また、該目標積算吸入空気量qair/cmdに対応して、内燃機関１の燃焼室４に吸入されるべき制御サイクル毎の吸入空気量、すなわち前記目標吸入空気量gair/cmdは、標準開度指令値θ0 と同じような時間的変化の形態でＦＩＲＥ経過時間t/fireに応じて変化するものとなる。
【０１２３】
また、上記のような目標積算吸入空気量qair/cmdは、あらかじめタイムテーブルにより設定しておき、このタイムテーブルを用いて制御サイクル毎にＦＩＲＥ経過時間t/fireから求めるようにしてもよい。
【０１２４】
本実施形態のシステムでは、前記吸気量Ｆ／Ｂ制御補正処理は、基本的には、上記のように求めた前記推定積算吸入空気量qair/pre（これは触媒装置３に実際に与える熱量の積算値に相当する）を目標積算吸入空気量qair/cmd（これは触媒装置３に与えるべき熱量の積算値の目標値に相当する）に収束させるように（これらの偏差を解消するように）フィードバック制御処理によりバイパス開度の指令値の補正量を求め、この補正量により標準開度指令値θ0 を補正してバイパス開度の指令値を決定することで、前述した構造要因による吸入空気量のばらつきを補償し、ひいては、触媒装置３の昇温形態のばらつきを解消するものである。
【０１２５】
この場合、吸気量Ｆ／Ｂ制御補正処理では、後述の如く、推定積算吸入空気量qair/preと目標積算吸入空気量qair/cmdとの偏差の減衰速度（推定積算吸入空気量qair/preの目標積算吸入空気量qair/cmdへの収束速度）を状況によって可変的に設定することが好ましいこと等から、該減衰速度を任意に所望の減衰速度に指定可能な応答指定型制御としてのスライディングモード制御（より詳しくは適応スライディングモード制御）を上記のフィードバック制御処理に用いる。
【０１２６】
この適応スライディングモード制御（以下、吸気側適応ＳＬＤ制御という）を用いて、推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させるようにバイパス開度の指令値を補正する吸気量Ｆ／Ｂ制御補正処理のアルゴリズムは本実施形態では次のように構築されている。尚、以下の説明においては説明の便宜上、しばらくの間、上記推定積算吸入空気量qair/preを含めて内燃機関１の燃焼室４の実際の吸入空気量の積算値を参照符号Ｑａを用いて積算吸入空気量Ｑａと称し、その目標値を参照符号ｑを用いて目標積算量ｑ（これは、目標積算吸入空気量qair/cmdに相当する）と称する。また、バイパス開度の指令値を一般的に参照符号Θを用いて開度指令Θと称する。
【０１２７】
まず、内燃機関１の燃焼室４の前記制御サイクル毎の実際の吸入空気量を参照符号Ｇcyl を用いて表し、この吸入空気量Ｇcyl と前記開度指令Θとの相関関係を次式（６）のように１次遅れ系の離散系（離散時間系）モデル（１次の自己回帰モデル）で表現する。
【０１２８】
【数６】

【０１２９】
ここで、式（６）中のα，βは、大気圧Ｐａや大気温度Ｔａ、吸気圧Ｐｂ、回転数Ｎｅ等に応じたモデルパラメータである。
【０１３０】
また、制御サイクル毎の前記積算吸入空気量Ｑａは、次式（７）により表されるので、
【０１３１】
【数７】

【０１３２】
この式（７）と前記式（６）とを用いて次式（８）が得られる。
【０１３３】
【数８】

【０１３４】
ここで、この式（８）中のＧcyl(k)は、式（７）によってＧcyl(k)＝Ｑa(k)−Ｑa(k-1)であるので、これを式（８）に代入して整理すると、次式（９）が得られる。
【０１３５】
【数９】

【０１３６】
この式（９）は、本実施形態のシステムにおいて、開度指令Θから積算吸入空気量Ｑａを生成する系、すなわち、前記吸気側適応ＳＬＤ制御の制御対象とすべき系を離散系のモデル（２次の自己回帰モデル。以下、吸気制御対象モデルという）で表現したものとなっている。すなわち、該吸気制御対象モデルは、吸気側適応ＳＬＤ制御の制御対象の出力としての各制御サイクルにおける積算吸入空気量Ｑa(k+1)を、それより過去の制御サイクルにおける積算吸入空気量Ｑａの時系列データＱa(k)，Ｑa(k-1)、並びに吸気側適応ＳＬＤ制御の制御対象の入力としての開度指令Θ(k) を用いて表現するものである。尚、式（９）において、積算吸入空気量Ｑa(k)，Ｑa(k-1)にそれぞれ係る係数ａ1 ，ａ2 と、開度指令Θ(k) に係る係数ｂ1 とは吸気制御対象モデルの挙動特性を規定するモデルパラメータである。
【０１３７】
本実施形態では、この吸気制御対象モデルに基づいて、吸気側適応ＳＬＤ制御のアルゴリズムを以下の如く構築している。
【０１３８】
すなわち、吸気側適応ＳＬＤ制御では、スライディングモード制御に必要な切換関数σ1 を、前記積算吸入空気量Ｑａと前記目標積算量ｑとの偏差Ｅｑ＝Ｑａ−ｑの制御サイクル毎の時系列データＥq(k)，Ｅq(k-1)を変数とする次式（１０）の線形関数により定義する。
【０１３９】
【数１０】

【０１４０】
ここで、式（１０）中のｓ1 ，ｓ2 は切換関数σ1 の各項の係数パラメータであり、次式（１１）の条件を満たすように設定する。
【０１４１】
【数１１】

【０１４２】
尚、本実施形態では、簡略化のためにｓ1 ＝１としている。また、本実施形態では、係数パラメータｓ2 の値（より一般的には、ｓ2 ／ｓ1 の値）を可変的に設定するのであるが、これについては後述する。
【０１４３】
このように切換関数σ1 を定義したとき、前記偏差Ｅｑの時系列データＥq(k)，Ｅq(k-1)の組から成る状態量（Ｅq(k)，Ｅq(k-1)）を図８に示す如く、σ1 ＝０なる関係式によって定義される切換線（これはすべり線とも言われる）上に収束させ、その収束状態を維持すると、状態量（Ｅq(k)，Ｅq(k-1)）を、外乱等の影響によらずに極めて安定に切換線σ1 ＝０上の平衡点、すなわち、Ｅq(k)＝Ｅq(k-1)＝０となる点に収束させることができる。
【０１４４】
尚、本実施形態では、切換関数σ1 に関する位相空間が２次元であるため（状態量（Ｅq(k)，Ｅq(k-1)）の成分が二つ）、切換線σ1 ＝０は直線となるが、位相空間が３次元である場合には、切換線は平面となり、この場合には、該切換線はすべり面といわれることもある。さらに、位相空間が４次元以上の高次元になると、切換線は幾何学的に図示できない超平面となる。
【０１４５】
一方、前記積算吸入空気量Ｑａを前記目標積算量ｑに収束させるために、前記式（９）によりモデル化した制御対象に与えるべき入力として吸気側適応ＳＬＤ制御が生成する制御入力、すなわち開度指令Θは、基本的には、前記状態量（Ｅq(k)，Ｅq(k-1)）を前述の如く定めた切換線σ1 ＝０に拘束するための制御則に基づいて定める等価制御入力Θeqと、状態量（Ｅq(k)，Ｅq(k-1)）を切換線σ1 ＝０に収束させるための制御則である到達則に基づいて定める到達則入力Θrch と、状態量（Ｅq(k)，Ｅq(k-1)）の切換線σ1 ＝０への収束に際しての外乱等の影響を排除するための制御則である適応則に基づいて定める適応則入力Θadp との総和である（次式（１２）を参照）。
【０１４６】
【数１２】

【０１４７】
尚、通常のスライディングモード制御では、適応則を考慮せず、この場合には、適応則入力Θadp は省略される。
【０１４８】
この場合、等価制御入力Θeqは、次式（１３）により与えられる。
【０１４９】
【数１３】

【０１５０】
尚、この式（１３）は、状態量（Ｅq(k)，Ｅq(k-1)）を切換線σ1 ＝０に拘束するためのσ1(k)＝σ1(k-1)なる条件と、前記吸気制御対象モデルの式（９）とに基づいて導出することができるものである。
【０１５１】
また、到達則入力Θrch 及び適応則入力Θadp は、それらを定める種々の手法が考えられるが、本実施形態では、到達則入力Θrch 及び適応則入力Θadp は、それぞれ前記切換関数σ1 の値、及び該切換関数σ1 の値の積算値（積分値）に比例させたものとし、それぞれ次式（１４）、（１５）により定める。
【０１５２】
【数１４】

【０１５３】
【数１５】

【０１５４】
ここで、式（１４）中のＦ1 は、到達則に係わるゲインを規定する係数であり、次式（１６）の条件を満たすように設定すればよい。但し、切換関数σ1 の値の切換線σ1 ＝０への収束に際してのチャタリングを低減する上では、式（１６）’の条件を満たすように係数Ｆ1 を設定することが好ましい。
【０１５５】
【数１６】

【０１５６】
また、式（１５）中のＦ2 は、適応則に係わるゲインを規定する係数であり、次式（１７）の条件を満たすように設定すればよい。尚、式（１７）中のΔＴは制御サイクル（制御周期）である。
【０１５７】
【数１７】

【０１５８】
本実施形態で用いる吸気側適応ＳＬＤ制御のアルゴリズムでは、基本的には、式（１３）〜（１５）により等価制御入力Θeq、到達則入力Θrch 及び適応則入力Θadp を求め、それらの総和を開度指令Θとして生成することで、前記積算吸入空気量Ｑａを前記目標積算量ｑに収束させるようにＦＩＲＥモードの動作中の吸入空気量を操作することが可能である。
【０１５９】
ところで、等価制御入力Θeq、到達則入力Θrch 及び適応則入力Θadp をそれぞれ前記式（１３）〜（１５）により求めるためには、前記式（９）により表した吸気制御対象モデルのモデルパラメータａ1 ，ａ2 ，ｂ1 の値を同定しておく必要がある。しかるに、これらのモデルパラメータａ1 ，ａ2 ，ｂ1 の値は、ＦＩＲＥモードの動作中の種々様々の要因の影響を受け、それらの値の最適な同定を行うことは煩雑なものとなりやすい。
【０１６０】
このため、本実施形態では、次のようにモデルパラメータａ1 ，ａ2 ，ｂ1 を排除し、吸気側適応ＳＬＤ制御の簡略化したアルゴリズムを構築する。
【０１６１】
まず、到達則入力Θrch 及び適応則入力Θadp に関しては、これらを求めるための前記式（１４）、（１５）中に含まれるモデルパラメータはｂ1 だけである。そして、式（１４）で（Ｆ1 ／ｂ1 ）を改めてＦｘとおき、また、式（１５）で（Ｆ2 ／ｂ1 ）を改めてＦｙとおけば、式（１４）、（１５）はそれぞれ次式（１８）、（１９）に書き換えられる。
【０１６２】
【数１８】

【０１６３】
【数１９】

【０１６４】
従って、これらの式（１８）、（１９）によって、モデルパラメータｂ1 を用いずに到達則入力Θrch 及び適応則入力Θadp を求めることができる。
【０１６５】
尚、この場合、式（１８）、（１９）中の係数Ｆｘ，Ｆｙの値は、切換関数σ1 の値の切換線σ1 ＝０への収束の安定性や速応性等を考慮して、実験やシミュレーションを通じて定めればよい。
【０１６６】
次に、等価制御入力Θeqに関しては、これを求めるための前記式（１３）は、前記偏差Ｅｑ＝Ｑａ−ｑを用いて次式（２０）に書き換えることができる。
【０１６７】
【数２０】

【０１６８】
この式（２０）中の第１番目の大括弧を含む項は、積算吸入空気量Ｑａと目標積算量ｑとの偏差Ｅｑに基づくフィードバック項で、第２番目の大括弧を含む項は、目標積算量ｑのみに基づくフィードフォワード項である。以下、このフィードバック項及びフィードフォワード項をそれぞれ次式（２１）、（２２）のようにΘeq/fb 、Θeq/ff とおく。
【０１６９】
【数２１】

【０１７０】
【数２２】

【０１７１】
ここで、上記フィードフォワード項Θeq/ff は、前記偏差Ｅｑが定常的に「０」となるような状態で制御対象に与えるべき入力（開度指令Θ）である。一方、前記標準開度指令値θ0 は、本実施形態では目標積算量ｑを規定するものであると共に、該標準開度指令値θ0 に対して吸入空気量、ひいては積算吸入空気量がその目標とすべき値に一義的に定まりるものとしてフィードフォワード的に設定したものである。
【０１７２】
従って、式（２０）中のフィードフォワード項Θeq/ff は、モデルパラメータａ1 ，ａ2 ，ｂ1 を含まない標準開度指令値θ0 に置き換えることができる。
【０１７３】
さらに、上記フィードバック項Θeq/fb に関しては、このフィードバック項Θeq/fb を含む等価制御入力Θeqは、前記状態量（Ｅq(k)，Ｅq(k-1)）を前記切換線σ1 ＝０上に拘束するための制御入力であるが、本願発明者等の各種検討によれば、本実施形態のシステムでは、状態量（Ｅq(k)，Ｅq(k-1)）が切換線σ1 ＝０の近傍に存する状態において安定性が高い。また、本実施形態では、前記適応則入力Θadp を採用することで、状態量（Ｅq(k)，Ｅq(k-1)）の切換線σ1 ＝０上への収束の安定性を高めることができる。
【０１７４】
このため、本実施形態のシステムでは、上記フィードバック項Θeq/fb を省略しても実用上、制御性が損なわれることはないと考えられる。
【０１７５】
このようなことを考慮すると、等価制御入力Θeqは、そのフィードバック項Θeq/fb を省略し、フィードフォワード項Θeq/ff を標準開度指令値θ0 で置き換えたものとしてもよいと考えられ、このようにすれば、等価制御入力Θeqは、モデルパラメータａ1 ，ａ2 ，ｂ1 を用いずに求めることができる。
【０１７６】
そこで、本実施形態では、吸気側適応ＳＬＤ制御の等価制御入力Θeqは、次式（２３）により与えるものとする。
【０１７７】
【数２３】

【０１７８】
以上のことから、本実施形態における吸気量Ｆ／Ｂ制御補正処理では、前述した構造要因による吸入空気量のばらつきを補償するために前記吸気側適応ＳＬＤ制御により決定する制御対象への入力、すなわち開度指令Θを次式（２４）の演算により求める。
【０１７９】
【数２４】

【０１８０】
別の言い方をすれば、本実施形態では、前記式（１８）、（１９）により求まる到達則入力Θrch 及び適応則入力Θadp の総和（＝Θrch ＋Θadp ）を次式（２５）のようにバイパス開度の指令値の補正量i/sld （以下、ＳＬＤ開度補正量i/sld という）として求め、このＳＬＤ開度補正量i/sld により標準開度指令値θ0 を補正する（標準開度指令値θ0 にＳＬＤ開度補正量i/sld を加算する）ことで、構造要因によるばらつきを補償するための開度指令Θを求める。
【０１８１】
【数２５】

【０１８２】
この場合において、ＳＬＤ開度補正量i/sld （＝Θrch ＋Θadp ）を求めるために必要となる切換関数σ1 の値は、本実施形態では、前記式（１０）の積算吸入空気量Ｑａとして前記式（２）により求まる推定積算吸入空気量qair/preを用い、且つ、目標積算量ｑとして前記式（５）により求まる目標積算吸入空気量qair/cmdを用いた次式（２６）の演算により求める。
【０１８３】
【数２６】

【０１８４】
尚、本実施形態では、ＦＩＲＥモードの中断動作中（前記ＦＩＲＥ中断フラグf/fpauseが「１」に設定されている状態）は、ＳＬＤ開度補正量i/sld の算出は中断する（ＳＬＤ開度補正量i/sld の値を中断動作の直前の値に保持する）が、推定積算吸入空気量qair/preと目標積算吸入空気量qair/cmdの算出は継続する（但し、推定積算吸入空気量qair/preに関しては内燃機関１のフュエルカット中を除く）。そして、前記ＦＩＲＥ制限時間TFIRELMTが経過する前に、中断動作を解除する状態となった場合には、ＳＬＤ開度補正量i/sld の算出及びそれに応じた標準開度指令値θ0 の補正を再開する。
【０１８５】
このようにするのは次の理由による。すなわち、ＦＩＲＥモードの中断動作中は、基本的には、車両のアクセルペダルの操作によって、車両の走行、もしくは内燃機関１のからぶかしを行う状況である。そして、この状況では、前記スロットル弁５がアクセル操作量Ａｐに応じた開度に開かれるので、内燃機関１の燃焼室４の実際の吸入空気量は、バイパス弁７を介した吸入空気量に、スロットル弁５を介した吸入空気量を加えたものとなる。
【０１８６】
このような状況では、推定積算吸入空気量qair/preは、バイパス弁７とスロットル弁５との両者による吸入空気量の積算値となるので、これをバイパス弁７の標準開度指令値θ0 に応じて定まる目標積算吸入空気量qair/cmdに収束させるようにバイパス開度を操作することは、アクセルペダルの操作に応じた内燃機関１の動作性能を確保する上で好ましくない。このために、本実施形態では、ＦＩＲＥモードの中断動作中は、ＳＬＤ開度補正量i/sld の算出を中断する。
【０１８７】
また、ＦＩＲＥモードの中断動作中は、内燃機関１の燃焼室４の実際の吸入空気量は、バイパス弁７を介した吸入空気量に、スロットル弁５を介した吸入空気量を加えたものとなるので、触媒装置３に与えられる熱量がさらに多くなる。従って、ＦＩＲＥモードの中断動作中に触媒装置３が十分に昇温・活性化される場合もあるが、該中断動作が短時間で解除されるような場合も多く、この場合には、該触媒装置３の昇温・活性化が未だ不十分なものとなることがある。このために、本実施形態では、ＦＩＲＥモードの中断動作中も推定積算吸入空気量qair/preと目標積算吸入空気量qair/cmdの算出を継続して行っておき、該中断動作の解除後にＳＬＤ開度補正量i/sld の算出を再開して標準開度指令値θ0 の補正を行うことで、ＦＩＲＥモードの動作中における触媒装置３の昇温・活性化を確実なものとする。但し、この場合において、推定積算吸入空気量qair/preに関しては、内燃機関１のフュエルカット中は、燃焼室４に吸入される空気が触媒装置３に与える熱量に寄与しないので、該フュエルカット中は、推定積算吸入空気量qair/preの算出を行わない。
【０１８８】
また、本実施形態では、後述する点火時期応動補正処理を行っている際にも、ＳＬＤ開度補正量i/sld の算出を中断する（ＳＬＤ開度補正量i/sld の値を中断動作の直前の値に保持する）。これは次の理由による。すなわち、点火時期応動補正処理は、詳細は後述するが、開度指令Θを標準開度指令値θ0 に対してフィードフォワード的に減少側に補正する処理であり、このような処理を行っている際に、ＳＬＤ開度補正量i/sld の算出を行うと、点火時期応動補正処理による開度指令Θの減少分を打ち消すように、ＳＬＤ開度補正量i/sld が算出されてしまうからである。
【０１８９】
以上説明した内容が、吸気量Ｆ／Ｂ制御補正処理の基本的内容である。
【０１９０】
ところで、本実施形態において、内燃機関１の吸入空気量の増量を開始した直後の段階、すなわち、バイパス開度を上昇させていく段階は、内燃機関１の始動直後の状態であるため、このような段階で標準開度指令値θ0 をＳＬＤ開度補正量i/sld により大きく補正すると、内燃機関１の燃焼室４での混合気の燃焼状態の悪化や内燃機関１のエミッション状態の悪化を招く虞れがある。また、本実施形態では、目標積算吸入空気量qair/cmdが、内燃機関１の定常的な吸気状態を前提としているため、吸入空気量の増量の開始直後の段階では、目標積算吸入空気量qair/cmdの信頼性が乏しいと考える。このため、吸入空気量の増量の開始直後の段階では、目標積算吸入空気量qair/cmdと前記推定積算吸入空気量qair/preとの偏差Ｅｑが大きなものとなって、前記ＳＬＤ開度補正量i/sld も大きなものとなる虞れがある。
【０１９１】
このようなことを考慮し、本実施形態における吸気量Ｆ／Ｂ制御補正処理では、吸入空気量の増量の開始後（ＦＩＲＥモードの開始後）、所定時間TISLDLMT（ＳＬＤ補正制限時間TISLDLMTという。図７を参照）が経過するまで（ＦＩＲＥ経過時間t/fire≧TISLDLMTとなるまで）の間は、目標積算吸入空気量qair/cmd及び推定積算吸入空気量qair/preの値を強制的に「０」とする（このときＥｑ＝０となる）。このようにすることで、ＦＩＲＥ経過時間t/fireがＳＬＤ補正制限時間TISLDLMTに達するまでの内燃機関１の始動直後の状態では、ＳＬＤ開度補正量i/sld の値も「０」に維持し、該ＳＬＤ開度補正量i/sld による標準開度指令値θ0 の補正を行わないようにしている。
【０１９２】
さらに、本実施形態における吸気量Ｆ／Ｂ制御補正処理では、ＳＬＤ開度補正量i/sld による標準開度指令値θ0 の実際の補正を開始する直後においても、その補正を急激に行うと、内燃機関１の燃焼状態やエミッション性能を損なう虞れがある。このため、本実施形態では、吸気量Ｆ／Ｂ制御補正処理で用いる前記吸気側適応ＳＬＤ制御の応答指定特性を利用し、推定積算吸入空気量qair/preと目標積算吸入空気量qair/cmdとの偏差Ｅｑ（以下、吸気偏差Ｅｑという）の減衰速度（推定積算吸入空気量qair/preの目標積算吸入空気量qair/cmdへの収束速度）を、次のように可変的に設定するようにしている。
【０１９３】
すなわち、吸気側適応ＳＬＤ制御において、吸気偏差Ｅｑに係わる状態量（Ｅq(k)，Ｅq(k-1)）が前記切換線σ1 ＝０に収束した状態では、前記式（１０）から明らかなように、次式（２７）の関係式が成立する。
【０１９４】
【数２７】

【０１９５】
従って、切換関数σ1 の係数パラメータｓ1 ，ｓ2 の比（ｓ2 ／ｓ1 ）（但し、−１＜ｓ2 ／ｓ1 ＜１）の値は吸気偏差Ｅｑの「０」への減衰速度を規定するものとなる（｜ｓ2 ／ｓ1 ｜が「０」に近づく程、減衰速度は速くなる）。従って、この比（ｓ2 ／ｓ1 ）の値によって、吸気偏差Ｅｑの減衰速度を指定することができることとなり、これが、吸気側適応ＳＬＤ制御の応答指定特性である。
【０１９６】
尚、吸気偏差Ｅｑの減衰速度を速くするということは、吸気側適応ＳＬＤ制御によるフィードバック制御のゲインを大きくするということと同等である。また、式（２７）は、入力の無い一次遅れ系を表現しており、上記比（ｓ2 ／ｓ1 ）は、この一次遅れ系の極に相当するものである（以下、上記比（ｓ2 ／ｓ1 ）をポールpole/iと称する）。また、本実施形態では、ｓ1 ＝１に設定しており、この場合、pole/i＝ｓ2 である。また、吸気偏差Ｅｑの「０」への減衰は、非振動的であることが好ましく、このため、本実施形態では、ｓ2 ／ｓ1 ＝pole/i＜０としている（ｓ2 ／ｓ1 ＞０とすると、式（２７）から明らかなように吸気偏差Ｅｑの「０」への減衰は振動的になる）。
【０１９７】
本実施形態における吸気量Ｆ／Ｂ制御補正処理では、このような吸気側適応ＳＬＤ制御の応答指定特性を利用し、基本的には、ＦＩＲＥ経過時間t/fireから、図９に示すようにあらかじめ定めたデータテーブル（タイムテーブル）に基づいて制御サイクル毎に求まる値pole/itbl （以下、ポールテーブル値pole/itbl という）を前記ポールpole/iの値として設定することで、該ポールpole/iの値をＦＩＲＥ経過時間t/fireに応じて可変的に設定する。
【０１９８】
ここで、図９のデータテーブルでは、ＦＩＲＥ経過時間t/fireが所定値TPOLEVST（但し、TPOLEVST≧ＳＬＤ補正制限時間TISLDLMT）に達してから、ポールテーブル値pole/itbl 、ひいてはポールpole/iをＦＩＲＥ経過時間t/fireの増加に伴い、所定の下限値pole/i0 （＜０。本実施形態では「−１」）から所定の定常値pole/ix （pole/i0 ＜pole/ix ＜０）に向かって徐々に増加させ（ポールテーブル値pole/itbl の絶対値を徐々に小さくしていく）、定常値pole/ix に達した後（ＦＩＲＥ経過時間t/fireが図９中の所定値TPOLEXに達した後）は、該定常値pole/ix に維持するようにしている。これにより、本実施形態では、ＳＬＤ開度補正量i/sld による標準開度指令値θ0 の補正を開始する直後は、吸気偏差Ｅｑの減衰速度を徐々に早めていく（推定積算吸入空気量qair/preの目標積算吸入空気量qair/cmdへの収束を緩やかに行う）ようにしていると共に、ＦＩＲＥ経過時間t/fireが所定値TPOLEXに達するまでは、該所定値TPOLEXに達した後よりも吸気偏差Ｅｑの減衰速度を遅くするようにしている。
【０１９９】
尚、ポールテーブル値pole/itbl によるポールpole/iの上記のような増加は、基本的には内燃機関１の始動後の初期段階（吸入空気量を増加させていく段階）で行う。
【０２００】
また、本実施形態では、ＦＩＲＥモードの中断動作中は、吸気量Ｆ／Ｂ制御補正処理（より正確には、ＳＬＤ開度補正量i/sld の算出処理）を行わないようにしており、この状態では、ポールpole/iをポールテーブル値pole/itbl の前記下限値pole/i0 に設定するようにしている。そして、ＦＩＲＥモードの中断動作が解除されたとき、ポールpole/iを上記下限値pole/i0 からポールテーブル値pole/itbl に向かって徐々に復帰させる（図９の仮想線を参照）ようにしている。
【０２０１】
次に、前述した大気条件による吸入空気量のばらつきの影響を補償するための前記大気条件補正処理について説明する。
【０２０２】
まず、前記大気条件のうちの大気圧に関して説明する。尚、ここでの説明では、大気温度は一定であり、また、バイパス開度は開度指令Θに等しいものとする。
【０２０３】
内燃機関１の燃焼室４内の圧力をＰcyl 、吸気バルブ１１（図２参照）の開度（有効開口面積）をＡcyl とおくと、該燃焼室４の吸入空気量Ｇcyl は、これらの圧力Ｐcyl 、開度Ａcyl 及び吸気圧Ｐｂとを用いて、一般的に次式（２８）により表される。
【０２０４】
【数２８】

【０２０５】
ここで、Ｃｉは、前記式（３）に関して説明した如く、空気密度に応じた係数である。
【０２０６】
この場合、内燃機関１では、式（２８）中における燃焼室４内の圧力をPcyl、吸気バルブ１１の開度Ａcyl は基本的には一定である。また、係数Ｃｉも大気温度Ｔａが一定であれば一定と考えてよい。
【０２０７】
従って、式（２８）から、燃焼室４内の吸入空気量Ｇcyl を大気圧によらないものとするためには、吸気圧Ｐｂが大気圧Ｐａに応じて変化しないことが必要となることが判る。
【０２０８】
一方、バイパス開度の開度指令Θに対して、バイパス弁７を通る空気量Ｇｉは、前記式（３）と同様、次式（２９）により表される。
【０２０９】
【数２９】

【０２１０】
さらに、大気圧Ｐａが所定の標準大気圧（以下、これに参照符号Ｐa0を付する）である状態において、開度指令Θを前記標準開度指令値θ0 （これは前述の通り標準大気圧Ｐa0を前提として定めたものである）としたときにバイパス弁７を通る空気量をＧi0（これは前記目標吸入空気量gair/cmdに相当する）とおくと、該空気量Ｇi0（以下、標準空気量Ｇi0という）は、次式（３０）により表される。
【０２１１】
【数３０】

【０２１２】
ここで、式（３０）中のＰb0は、標準大気圧Ｐa0において、バイパス開度を標準開度指令値θ0 としたときにチャンバー１３（図２参照）内に発生する吸気圧（以下、標準吸気圧Ｐb0という）である。
【０２１３】
また、周知の気体の状態方程式によれば、チャンバー１３内の吸気圧Ｐｂが変動しないための条件は、チャンバー１３内に流入する空気量、すなわちバイパス弁７を通る空気量Ｇi と、チャンバー１３から流出する空気量、すなわち燃焼室４の吸入空気量Ｇcyl とが等しくなることである。
【０２１４】
以上のことから、大気圧Ｐａが標準大気圧Ｐa0に対して変動したとき、燃焼室４の吸入空気量Ｇcyl が変化しないようにするための開度指令Θは、式（２９）中の吸気圧Ｐｂが、前記標準吸気圧Ｐb0と等しくなり、且つ、式（２９）により表される空気量Ｇｉが前記標準空気量Ｇi0（式（３０））と等しくなるように定めればよい。
【０２１５】
すなわち、その開度指令Θは、次式（３１）の条件を満たすように定めればよい。
【０２１６】
【数３１】

【０２１７】
そして、この式（３１）を開度指令Θについて解くと、次式（３２）が得られる。
【０２１８】
【数３２】

【０２１９】
従って、基本的には、この式（３２）に基づいて標準開度指令値θ0 を補正して開度指令Θを決定すれば、大気圧による吸入空気量のばらつきを補償しすることができる。つまり、式（３２）中の平方根の値kpa （以下、大気圧補正係数kpa という）を開度指令に乗算して補正することで、大気圧による吸入空気量のばらつきを補償するための開度指令を決定することができる。
【０２２０】
ところで、本実施形態では、標準開度指令値θ0 を前述の如く時間的に変化させるため、式（３２）の演算で使用する標準吸気圧Ｐb0も変化し、従って、大気圧による吸入空気量のばらつきを的確に補償する上では、標準吸気圧Ｐb0の値をあらかじめ定めたデータテーブル等を用いて標準開度指令値θ0 に応じて適宜変更することが好ましいと考えられる。但し、本実施形態では、ＦＩＲＥモードにおける定常的な内燃機関１の運転中は、実際上、吸気圧Ｐｂの変動は小さいということ、並びに制御系の安定性を考慮し、式（３２）中の標準吸気圧Ｐb0としてあらかじめ定めた所定値（固定値）を用いる。
【０２２１】
また、本実施形態では、コントローラ２の演算負荷を軽減するために、実際には、式（３２）の演算を直接的には行わず、次のような処理を行う。
【０２２２】
すなわち、制御サイクル毎に、あらかじめ定めた標準大気圧Ｐa0と、標準吸気圧Ｐb0と、内燃機関１の始動時（始動モード）において前記大気圧センサ１８により検出される大気圧Ｐａとから次式（３３）により定義するパラメータratio/dpa （以下、大気圧補正用パラメータratio/dpa という）、すなわち、式（３２）の平方根（√）内の値を求める。
【０２２３】
【数３３】

【０２２４】
そして、あらかじめ大気圧補正用パラメータratio/dpa の種々の値に対して平方根の演算を行って定めた図１０のデータテーブルを用意しておき、先に求めた大気圧補正用パラメータratio/dpa の平方根を前記大気圧補正係数kpa として求める。そして、この大気圧補正係数kpa を用いて開度指令を補正（乗算補正）する。
【０２２５】
尚、この大気圧補正係数kpa は、大気圧センサ１８により検出される大気圧Ｐａが標準大気圧Ｐa0であれば、「１」であり、該大気圧Ｐａが大きくなる程、値が小さくなる。
【０２２６】
以上説明した内容が、前記大気条件補正処理において、大気圧による吸入空気量のばらつきを補償するための処理の基本的内容である。
【０２２７】
次に、大気条件のうちの大気温度に関して説明する。前記式（２８）から明らかなように内燃機関１の燃焼室４の吸入空気量Ｇcyl は、大気密度に応じた係数Ｃｉの影響を受け、大気密度が高い程、多くなる。そして、大気密度は、大気温度が高い程、低くなるので、燃焼室４の吸入空気量Ｇcyl は大気温度が高い程、少なくなる。
【０２２８】
従って、大気温度が変化しても、燃焼室４の吸入空気量Ｇcyl を変化させないようにするためには、大気温度が高い程、バイパス開度が大きくなるように開度指令Θを補正してやればよい。
【０２２９】
そこで、本実施形態では、内燃機関１の始動時（始動モード）において前記大気温度センサ１７により検出される大気温度Ｔａから、図１１に示す如くあらかじめ実験等に基づき定めたデータテーブルに基づいて補正係数kta （以下、大気温度補正係数kta という）を求め、この大気温度補正係数kta を用いて開度指令を補正（乗算補正）する。
【０２３０】
この場合、本実施形態では、標準大気温度Ｔa0（例えば２５°Ｃ）における吸入空気量を基準としているので、大気温度補正係数kta は、検出される大気温度Ｔａが標準大気温度Ｔa0であるとき「１」であり、大気温度Ｔａが高くなる程、値が大きくなる。
【０２３１】
これが、前記大気条件補正処理において、大気温度による吸入空気量のばらつきを補償するための処理の基本的内容である。
【０２３２】
次に、前記学習補正処理について説明する。
【０２３３】
本実施形態では、前述した如く、前記構造要因による吸入空気量のばらつきを補償するために、吸気側適応ＳＬＤ制御を用いた吸気量Ｆ／Ｂ制御補正処理によって、制御サイクル毎に前記ＳＬＤ開度補正量i/sld を求め、このＳＬＤ開度補正量i/sld によって標準開度指令値θ0 を補正する。この場合、バイパス弁７の経年劣化等により、標準開度指令値θ0 に対する実際の吸入空気量が、該標準開度指令値θ0 に対応した本来の標準的な吸入空気量に対してばらつきが比較的大きくなっているような状態では、前記積算吸入空気量Ｑａ（推定積算吸入空気量qair/pre）が目標積算量ｑ（目標積算吸入空気量qair/cmd）に対して未収束の段階でのＳＬＤ開度補正量i/sld は大きなものとなり、また、その時間的変化も大きなものとなる。このため、ＳＬＤ開度補正量i/sld により標準開度指令値θ0 を補正して決定した開度指令Θ、ひいては実際の吸入空気量の時間的変化のパターンは、ＦＩＲＥモードの初期段階において、標準開度指令値θ0 の時間的変化のパターン（目標とする吸入空気量の時間的変化のパターン）に対して大きく逸脱した時間的変化を生じるものとなることがある。
【０２３４】
しかるに、特にＦＩＲＥモードの初期段階（内燃機関１の始動後まもなくの状態）では、燃焼室４での混合気の燃焼状態が不安定なものとなりやすく、開度指令Θの時間的変化のパターンが標準開度指令値θ0 の時間的変化のパターンに対して大きく逸脱した変化を呈すると、該燃焼状態の悪化や、エミッション状態の悪化を引き起こす虞れがある。
【０２３５】
そこで、本実施形態における学習補正処理では、ＦＩＲＥモードの動作中に制御サイクル毎に求めるを学習し、次回のＦＩＲＥモードの動作の全期間にわたって、標準開度指令値θ0 を乗算補正するための補正係数kilearn （以下、学習補正係数kilearn という）を求める。そして、この学習補正係数kilearn を標準開度指令値θ0 に乗算することで、標準開度指令値θ0 の時間的変化のパターンと整合した安定な時間的変化を呈する開度指令Θを生成する。
【０２３６】
この場合、学習補正係数kilearn は、本実施形態では次のように決定する。
【０２３７】
すなわち、各回のＦＩＲＥモードの動作中に制御サイクル毎に求められるＳＬＤ開度補正量i/sld から、次式（３４）によって、該ＳＬＤ開度補正量i/sld に対応する実際の吸入空気量の補正量gair/sld（以下、ＳＬＤ吸気補正量gair/sldという）を制御サイクル毎に求める。
【０２３８】
【数３４】

【０２３９】
ここで、この式（３４）は、前記目標吸入空気量gair/cmd、すなわち標準開度指令値θ0 に対応した実際の吸入空気量（１ＴＤＣ当たり）を求めるための式（４）と同様の式であり、式（３４）中のＧa2は、式（４）中のＧa2と同一である。
【０２４０】
さらに、このＳＬＤ吸気補正量gair/sldを次式（３５）により制御サイクル毎に累積加算することで、ＳＬＤ吸気補正量gair/sldの積算値qair/sld（以下、ＳＬＤ積算吸気補正量qair/sldという）を求める。
【０２４１】
【数３５】

【０２４２】
そして、このＳＬＤ積算吸気補正量qair/sldの前記目標積算吸入空気量qair/cmdに対する比（qair/sld／qair/cmd）から次式（３６）により求まる値を前記学習補正係数kilearn の基本値vpskisld（以下、基本学習補正係数vpskisldという）を求める。
【０２４３】
【数３６】

【０２４４】
尚、この基本学習補正係数vpskisldは、基本的には、ＦＩＲＥモードの動作が終了するまで制御サイクル毎に算出するが、ＦＩＲＥモードの前記中断動作が行われる場合や、後述する点火時期応動補正処理が行われる場合にあっては、前述の如くＳＬＤ開度補正量i/sld の算出を中断するので、ＦＩＲＥモードの中断動作や点火時期応動補正処理が開始される前（前記学習演算終了フラグf/flrnend が「１」に設定される前）までで基本学習補正係数vpskisldの算出を終了する。
【０２４５】
そして、このようにして各回のＦＩＲＥモードの動作中に最終的に求められた基本学習補正係数vpskisldに対して次式（３７）によるなまし演算処理（フィルタリング処理）を施すことで、次回のＦＩＲＥモードの動作に際して標準開度指令値θ0 を補正するための学習補正係数kilearn を求める。
【０２４６】
【数３７】

【０２４７】
ここで、式（３７）中のkilearn(j)は、今回のＦＩＲＥモードの動作によって新たに決定する学習補正係数kilearn を意味し、kilearn(j-1)は、前回のＦＩＲＥモードの動作によって決定された学習補正係数kilearn を意味する。また、式（３７）中のＣkiは、あらかじめ定めた「１」以下の定数である。
【０２４８】
尚、この場合において、各回のＦＩＲＥモードの動作において、最終的な基本学習補正係数vpskisldが求められた時のＦＩＲＥ経過時間t/fire（これは、前記図５のフローチャートのＳＴＥＰ５−５，５−１６等で設定されるパラメータt/kil の値である）が所定時間に満たない場合は、その基本学習補正係数vpskisldは、学習補正係数kilearn を求める（更新する）ためには使用せず、現在の学習補正係数kilearn の値を維持する。これは、ＦＩＲＥ経過時間t/fireが短い段階で得られる基本学習補正係数vpskisldの信頼性が乏しいからである。
【０２４９】
以上説明した処理が、学習補正処理の基本的内容である。
【０２５０】
次に前記点火時期応動補正処理について説明する。
【０２５１】
本実施形態では、前述した如く、内燃機関１の暖機がある程度進行すると（ＦＩＲＥ経過時間t/fireがある程度大きくなると）、該内燃機関１の各部のフリクションの低下による内燃機関１の回転数Ｎｅの上昇傾向を抑え、点火時期が前記点火時期操作回転数Ｆ／Ｂ制御により過度に遅角側に操作されるのを予防するために、標準開度指令値θ0 を緩やかに減少させていくようにしている（図７を参照）。
【０２５２】
しかるに、内燃機関１の暖機の進行に伴うフリクションの低下の形態は、種々様々の要因の影響を受け、該フリクションの低下が予想以上に早期に始まったり、あるいは、そのフリクションの低下度合いが予想以上に大きなものとなることがある。
【０２５３】
そして、このような場合には、標準開度指令値θ0 を前記のように緩やかに減少させていっても、内燃機関１の回転数Ｎｅの上昇傾向を十分に抑えることができなくなる。その結果、点火時期操作回転数Ｆ／Ｂ制御によりに操作される点火時期が、実際に操作し得る遅角側の限界値まで達してしまい、ひいては、回転数Ｎｅを目標回転数ne/fire にフィードバック制御することができなくなってしまう。
【０２５４】
前記点火時期応動補正処理はこのような事態を回避するために行う処理であり、点火時期操作回転数Ｆ／Ｂ制御により後述の如く決定される点火時期の指令値が、遅角側の限界値よりも若干進角側に定めた所定の閾値を超えて遅角側の値になったときに、その状態が解消されるまで（点火時期の指令値が閾値以上の進角側の値に復帰するまで）、制御サイクル毎に、標準開度指令値θ0 を所定量づつ減少側に補正する。
【０２５５】
すなわち、例えばバイパス開度が図１２の上段に示すように標準開度指令値θ0 に従って操作されている状態で、図１２の下段のＡ12領域に示すように点火時期操作回転数Ｆ／Ｂ制御により決定される点火時期の指令値iglog が、遅角側の限界値IGLGG よりも若干大きい閾値IGX よりも遅角側に低下すると、開度指令Θを標準開度指令値θ0 に対して、図１２の上段に示すように、ある補正量分θdec （以下、点火時期応動開度補正量θdec という）、減少させる（Θ＝θ0 −θdec とする）。
【０２５６】
この場合、点火時期応動開度補正量θdec は、図１２の上段のＢ12領域に示すように、点火時期の指令値iglog が閾値IGX 以上の進角側に復帰するまで、制御サイクル毎に所定量Δθdec （＞０。以下、開度減少単位量Δθdec という）づつ、増加させる（図１５のＳＴＥＰ１５−９の式を参照）。
【０２５７】
そして、図１２の下段のＣ12領域に示すように点火時期の指令値iglog が閾値IGX 以上の進角側に復帰した後は、図１２の上段のＤ12領域に示すように点火時期応動開度補正量θdec を上記復帰時点の値に保持し（点火時期応動開度補正量θdec を前記開度減少単位量Δθdec づつ増加させる処理を中止する）、その保持した点火時期応動開度補正量θdec だけ、開度指令Θを標準開度指令値θ0 よりも減少側に補正する。
【０２５８】
尚、本実施形態では、点火時期応動開度補正量θdec を制御サイクル毎に増加させる前記開度減少単位量Δθdec は、内燃機関１の始動時における機関温度Ｔｗから、図１３に示す如くあらかじめ定めたデータテーブルに基づいて決定する。この場合、特に、内燃機関１の始動時の機関温度Ｔｗが高温領域である場合に、該機関温度Ｔｗが低中温領域である場合よりもフリクションの低下が大きく生じ易いことから、図１３のデータテーブルでは、機関温度Ｔｗの高温領域における開度減少単位量Δθdec を低中温領域よりも大きくするように設定している。
【０２５９】
以上説明した内容が、前記点火時期応動補正処理の基本的内容である。
【０２６０】
以上説明した内容をふまえて、前記図４のＳＴＥＰ４−５において前記吸入空気量制御段２５が行うバイパス開度の指令値（開度指令）θCMD の生成処理を次に具体的に説明する。
【０２６１】
図１４のフローチャートを参照して、前記ＳＴＥＰ４−５では、まず、前記ＳＴＥＰ４−４の条件判断処理（図１５参照）で前述の如く制御サイクル毎に設定されるＦＩＲＥ実行可否フラグf/fireonの値（現在の制御サイクルで設定された値）を判断する（ＳＴＥＰ１４−１）。
【０２６２】
このとき、f/fireon＝１であるとき、すなわち、動作モードがＦＩＲＥモードに設定されているときには、前記標準開度指令値生成処理によって標準開度指令値θ0 を求めるために用いる前記Ｎレンジ基本値ifiret及びＤレンジ補正値iatfire と、前記大気条件補正処理に用いる前記大気圧補正係数kpa 及び大気温度補正係数kta と、前記点火時期応動補正処理に用いる前記開度減少単位量Δθdec とを、それぞれに対応して前述の如く定めたデータテーブルを用いて求める（ＳＴＥＰ１４−２）。
【０２６３】
すなわち、前記始動モード処理（ＳＴＥＰ４−２）で取得した内燃機関１の始動時の機関温度Ｔｗから、前記図６のデータテーブルに基づいてＮレンジ基本値ifiret及びＤレンジ補正値iatfire を求める。
【０２６４】
また、始動モード処理で取得した内燃機関１の始動時の大気圧Ｐａと、あらかじめ定められた標準大気圧Ｐa0 及び標準吸気圧Ｐb0 とから、前記式（３３）によって、前記大気圧補正用パラメータratio/dpa を算出し、このパラメータratio/dpa から、図１０のデータテーブルに基づいて大気圧補正係数kpa （＝パラメータratio/dpa の平方根）を求める。
【０２６５】
さらに、始動モード処理で取得した内燃機関１の始動時の大気温度Ｔａから、図１１のデータテーブルに基づいて大気温度補正係数kta を求める。
【０２６６】
また、始動モード処理で取得した内燃機関１の始動時の機関温度Ｔｗから、図１３のデータテーブルに基づいて開度減少単位量Δθdec を求める。
【０２６７】
尚、ＳＴＥＰ１４−２におけるこれらの処理は、始動モード処理においてあらかじめ行っておくようにしてもよい。
【０２６８】
次いで、吸入空気量操作手段２５は、標準開度指令値θ0 に対して前記吸気量Ｆ／Ｂ制御補正処理と前記点火時期応動補正処理とを施してなる予備開度指令θi/fireを算出する処理を次のように行う（ＳＴＥＰ１４−３）。
【０２６９】
すなわち、図１５のフローチャートを参照して、図示しないセンサにより検出される前記自動変速機の現在（今回の制御サイクル）のシフト位置を判断し（ＳＴＥＰ１５−１）、該シフト位置がＮレンジである場合には、標準開度指令値θ0 の前記基本値i/ftblを前記ＳＴＥＰ１４−２で求めたＮレンジ基本値ifiretとする（ＳＴＥＰ１５−２）。また、現在のシフト位置がＤレンジである場合には、上記Ｎレンジ基本値ifiretに、ＳＴＥＰ１４−２で求めたＤレンジ補正値iatfire を加算した値を基本値i/ftblとする（ＳＴＥＰ１５−３）。
【０２７０】
次いで、現在のＦＩＲＥ経過時間t/fireから、前記図７のデータテーブルに基づいて今回の制御サイクルにおける前記補正係数km/fire を求め（ＳＴＥＰ１５−４）、この補正係数km/fire をＳＴＥＰ１５−２あるいは１５−３で決定した基本値i/ftblに乗算することで、標準開度指令値θ0 を求める（ＳＴＥＰ１５−５）。これによりＦＩＲＥモードにおける制御サイクル毎の標準開度指令値θ0 が決定される。
【０２７１】
次いで、吸入空気量操作手段２５は、前記吸気偏差Ｅｑを算出する処理を図１６のフローチャートに示すように行う（ＳＴＥＰ１５−６）。
【０２７２】
すなわち、図１６を参照して、まず、前記吸気圧センサ１６により検出される現在の吸気圧Ｐｂとあらかじめ定めた所定値Ｇa1とから前記式（１）により今回の制御サイクルのおける推定吸入空気量gair/pre（１ＴＤＣ当たりの吸入空気量の推定値）を求める（ＳＴＥＰ１６−１）。
【０２７３】
次いで、内燃機関１のフュエルカット中であるか否かを判断し（ＳＴＥＰ１６−２）、フュエルカット中でない場合には、さらに、現在のＦＩＲＥ経過時間t/fireが前記ＳＬＤ補正制限時間TISLDLMTに達したか否かを判断する（ＳＴＥＰ１６−３）。
【０２７４】
このとき、t/fire＜TISLDLMTである場合には、今回の制御サイクルにおける推定積算吸入空気量qair/pre(k) の値を強制的に「０」とする（ＳＴＥＰ１６−４）。また、t/fire≧TISLDLMTである場合には、前記式（２）により推定吸入空気量gair/preを累積加算して推定積算吸入空気量qair/pre(k) を求める（ＳＴＥＰ１６−５）。
【０２７５】
尚、ＳＴＥＰ１６−２でフュエルカット中である場合には、そのとき内燃機関１の燃焼室に吸入される空気は、触媒装置３に与える熱量には寄与しない（フュエルカット中は内燃機関１の燃焼室４での混合気の燃焼は行われない）ため、推定積算吸入空気量qair/pre(k) は現状の値に保持される（ＳＴＥＰ１６−６）。
【０２７６】
以上のようなＳＴＥＰ１６−１〜１６−６の処理により、ＦＩＲＥモードでは（前記中断動作中も含む）、その開始後、ＳＬＤ補正制限時間TISLDLMTが経過した時から、内燃機関１のフュエルカット中の場合を除いて、逐次、触媒装置３に実際に与えられる熱量の積算値に相当する推定積算吸入空気量qair/pre(k) が求められていくこととなる。
【０２７７】
そして、ＳＬＤ補正制限時間TISLDLMTが経過するまで、すなわち、内燃機関１の始動直後の状態では、推定積算吸入空気量qair/pre(k) の値が強制的に「０」に制限されることとなる。
【０２７８】
尚、このようにＳＬＤ補正制限時間TISLDLMTが経過するまで、推定積算吸入空気量qair/pre(k) の値を制限するためには、例えばＳＬＤ補正制限時間TISLDLMTが経過するまで、推定吸入空気量gair/preの値を強制的に「０」に制限し、その制限した推定吸入空気量gair/preの値を用いて前記式（２）の演算を行うようにしてもよい。このようにしても、推定積算吸入空気量qair/pre(k) の値は、ＳＬＤ補正制限時間TISLDLMTが経過するまでは「０」に制限されることとなる。
【０２７９】
このようにして推定積算吸入空気量qair/pre(k) を求めた後、吸入空気量操作手段２５は、前記ＳＴＥＰ１５−５で求めた標準開度指令値θ0 と、前記回転数センサ１４により検出される現在の回転数Ｎｅと、あらかじめ定められた所定値Ｇa2とから前記式（４）により今回の制御サイクルにおける目標吸入空気量gair/cmd（１ＴＤＣ当たりの吸入空気量の目標値）を求める（ＳＴＥＰ１６−７）。
【０２８０】
さらに、前記ＳＴＥＰ１６−３と同じ判断を行い（ＳＴＥＰ１６−８）、このとき、t/fire＜TISLDLMTである場合には、今回の制御サイクルにおける目標積算吸入空気量qair/cmd(k) の値を強制的に「０」とする（ＳＴＥＰ１６−９）。また、t/fire≧TISLDLMTである場合には、前記式（５）により制御サイクル毎の目標吸入空気量gair/preを累積加算して目標積算吸入空気量qair/cmd(k) を求める（ＳＴＥＰ１６−１０）。
【０２８１】
以上のようなＳＴＥＰ１６−７〜１６−１０の処理により、触媒装置３に実際に与えられる熱量の積算値の目標値に相当する目標積算吸入空気量qair/cmd(k) は、ＦＩＲＥモードの開始後、ＳＬＤ補正制限時間TISLDLMTが経過した時から、ＦＩＲＥモードの中断動作中を含めて、逐次求められていくこととなる。
【０２８２】
そして、ＳＬＤ補正制限時間TISLDLMTが経過するまでの内燃機関１の始動直後の状態では、推定積算吸入空気量qair/pre(k) と同様、目標積算吸入空気量qair/cmd(k) の値が強制的に「０」に制限されることとなる。
【０２８３】
尚、このようにＳＬＤ補正制限時間TISLDLMTが経過するまで、目標積算吸入空気量qair/cmd(k) の値を制限するためには、ＳＬＤ補正制限時間TISLDLMTが経過するまで、目標吸入空気量gair/cmdの値を強制的に「０」に制限し、その制限した目標吸入空気量gair/cmdの値を用いて前記式（５）の演算を行うようにしてもよい。
【０２８４】
このようにして、今回の制御サイクルにおける推定積算吸入空気量qair/pre(k) と、目標積算吸入空気量qair/cmd(k) とを求めた後には、吸入空気量操作手段２５は、それらの差（qair/pre(k) −qair/cmd(k) ）を演算することで、今回の制御サイクルにおける前記吸気偏差Ｅq(k)を求め（ＳＴＥＰ１６−１１）、図１５のフローチャートの処理に復帰する。
【０２８５】
図１５の説明に戻って、上記のように吸気偏差Ｅｑを求めた後、吸入空気量操作手段２５は、次に、フラグf/dec の値を判断する（ＳＴＥＰ１５−７）。このフラグf/dec は、前記点火時期応動補正処理に係わるフラグであって、後述する点火時期の指令値iglog の生成処理において点火時期の指令値iglog が前記閾値IGX （図１２参照）よりも遅角側の値であるときf/dec ＝１とされ、該指令値iglog が前記閾値IGX 以上の進角側の値であるときf/dec ＝０とされるものである（以下、フラグf/dec を点火時期判別フラグf/dec という）。尚、この点火時期判別フラグf/dec は前記始動モード処理（ＳＴＥＰ４−２）において「０」に初期化される。
【０２８６】
このとき、f/dec ＝０である場合（点火時期の指令値iglog が、閾値IGX 以上の進角側の状態）には、吸入空気量操作手段２５は、前記吸気量Ｆ／Ｂ制御補正処理に係わるＳＬＤ開度補正量i/sld を算出する処理を図１７のフローチャートに示すように行う（ＳＴＥＰ１５−８）。
【０２８７】
すなわち、まず、前記ＦＩＲＥ中断フラグf/fpauseの現在の値を判断する（ＳＴＥＰ１７−１）。このとき、f/fpause＝１である場合、すなわち、ＦＩＲＥモードの前述の中断動作を行う状態である場合には、前記吸気量Ｆ／Ｂ制御補正処理で吸気偏差Ｅｑの減衰速度を規定する前記ポールpole/iの値をあらかじめ定めた前記下限値pole/i0 （図９を参照）に設定（初期化）した上で（ＳＴＥＰ１７−２）、直ちに図１５のルーチン処理に復帰する。
【０２８８】
従って、ＦＩＲＥモードの中断動作中は、ＳＬＤ開度補正量i/sld は現状の値（中断動作の開始前の値）に保持される。
【０２８９】
尚、ＳＬＤ開度補正量i/sld 及びポールpole/iは、前記始動モード処理（ＳＴＥＰ４−２）においてそれぞれ「０」、「下限値pole/i0 」に初期化される。
【０２９０】
一方、ＳＴＥＰ１７−１の判断で、f/fpause＝０である場合、すなわち、ＦＩＲＥモードの通常的な動作を行う状態である場合には、吸入空気量操作手段２５は、現在のＦＩＲＥ経過時間t/fireから、図９のデータテーブルに基づいて今回の制御サイクルにおける前記ポールテーブル値pole/itbl を求める（ＳＴＥＰ１７−３）。
【０２９１】
次いで、基本的には、ＳＴＥＰ１７−３で求めたポールテーブル値pole/itbl を吸気偏差Ｅｑの減衰速度を規定するポールpole/iの値として設定するのであるが、ＦＩＲＥモードの中断動作中に前記下限値pole/i0 （本実施形態では「−１」）に設定されるポールpole/iを該中断動作の終了後に徐々にＦＩＲＥ経過時間t/fireに応じたポールテーブル値pole/itbl に復帰させるために、次のような処理を行う。
【０２９２】
すなわち、現在のポールpole/iの値pole/i(k-1) にあらかじめ定めた単位増分値ΔPOLE/I（＞０）を加算した値（pole/i(k-1) ＋ΔPOLE/I）をＳＴＥＰ１７−３で求めたポールテーブル値pole/itbl と比較する（ＳＴＥＰ１７−４）。そして、pole/i(k-1) ＋ΔPOLE/I≧pole/itbl である場合には、ＳＴＥＰ１７−３で求めたポールテーブル値pole/itbl を今回の制御サイクルにおけるポールpole/i(k) の値として設定し（ＳＴＥＰ１７−５）、pole/i(k-1) ＋ΔPOLE/I＜pole/itbl である場合には、pole/i(k-1) ＋ΔPOLE/Iの値を今回の制御サイクルにおけるポールpole/i(k) の値として設定する（ＳＴＥＰ１７−６）。
【０２９３】
この処理によりＦＩＲＥモードの中断動作が終了した後には、ポールpole/iが前記下限値pole/i0 から、ＦＩＲＥ経過時間t/fireに応じたポールテーブル値pole/itbl に前記単位増分値ΔPOLE/Iづつ、徐々に復帰することとなる。
【０２９４】
尚、内燃機関１の始動直後にＦＩＲＥモードの中断動作が行われない場合には、ポールpole/iは、基本的にはポールテーブル値pole/itbl が設定され、該ポールテーブル値pole/itbl と同じ形態でＦＩＲＥ経過時間t/fireの増加に伴い値が変化する（このようになるように上記単位増分値ΔPOLE/Iの値が設定されている）。
【０２９５】
このようにしてポールpole/iの値を設定した後、吸入空気量操作手段２５は、さらに、該ポールpole/iの値と、後に詳細を説明する点火時期操作回転数Ｆ／Ｂ制御において設定されるポールpole/ig （これは点火時期操作回転数Ｆ／Ｂ制御による内燃機関１の回転数Ｎｅと目標回転数ne/fire との偏差の減衰速度を規定するパラメータである）からあらかじめ定めた所定の微小量ΔPOLE/IG （＞０）を減算した値とを比較する（ＳＴＥＰ１７−７）。
【０２９６】
そして、このとき、pole/i＜pole/ig −ΔPOLE/IG であれば、そのまま次のＳＴＥＰ１７−９に進むが、pole/i≧pole/ig −ΔPOLE/IG である場合、ポールpole/iの値を強制的に（pole/ig −ΔPOLE/IG ）に設定し直す（ＳＴＥＰ１７−８）。つまり、ポールpole/iの値は、点火時期操作回転数Ｆ／Ｂ制御において後述の如く設定されるポールpole/ig （＜０）よりも常に小さい値に（より正確には、１＞｜pole/i｜＞｜pole/ig ｜＞０となるように）設定される。これは次の理由による。
【０２９７】
すなわち、本実施形態において推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させるように行う吸気側適応ＳＬＤ制御（フィードバック制御）と、内燃機関１の回転数Ｎｅを目標回転数ne/fire に収束させるように行う点火時期操作回転数Ｆ／Ｂ制御とは互いに独立的に行われるものである一方、両者の制御は、内燃機関１の回転数Ｎｅに影響を及ぼす制御である。また、一般に、吸気側適応ＳＬＤ制御に基づくバイパス開度の変化に対する吸入空気量の変化の応答性は、点火時期操作回転数Ｆ／Ｂ制御に基づく点火時期の変化に対する回転数Ｎｅの変化の応答性に比して遅い。このため、吸気側適応ＳＬＤ制御に係わる吸気偏差Ｅｑの減衰速度を、点火時期操作回転数Ｆ／Ｂ制御による内燃機関１の回転数Ｎｅと目標回転数ne/fire との偏差の減衰速度よりも速めるようにすると、両者の制御が互いに干渉して、内燃機関１の回転数Ｎｅが不安定なものとなる虞れがある。
【０２９８】
このために、本実施形態では、吸気側適応ＳＬＤ制御に係わるポールpole/iを、前記の如く｜pole/i｜＞｜pole/ig ｜となるように設定し、これにより、吸気側適応ＳＬＤ制御に係わる吸気偏差Ｅｑの減衰速度を、点火時期操作回転数Ｆ／Ｂ制御による内燃機関１の回転数Ｎｅと目標回転数ne/fire との偏差の減衰速度よりも遅くし、ひいては両制御が互いに干渉するのを回避する。
【０２９９】
次いで、ＳＴＥＰ１７−９では、ポールpole/iの値を「−１」と比較し、pole/i≦−１である場合（このような場合は、ＳＴＥＰ１７−８の処理によって生じることがある）には、ポールpole/iの値を強制的に「−１」に設定した上で（ＳＴＥＰ１７−１０）、ＳＴＥＰ１７−１１に進む。
【０３００】
そして、ＳＴＥＰ１７−１１では、上記のようにして決定したポールpole/iの値と、前記ＳＴＥＰ１５−６で前述の如く求めた今回の制御サイクルにおける吸気偏差Ｅq(k)及び前回の制御サイクルにおける吸気偏差Ｅq(k-1)とから前記式（１０）（詳しくは、式（１０）の係数パラメータｓ1 ，ｓ2 をそれぞれ「１」、「pole/i」で置き換えた式）により前記切換関数σ1 の値を求める。
【０３０１】
さらに、この切換関数σ1 の値を用いて、前記式（１８）、（１９）の演算を行うことで（この場合ｓ1 ＝１）、前記吸気側適応ＳＬＤ制御における到達則入力Θrch 及び適応則入力Θadp の値を求め（ＳＴＥＰ１７−１２）、この到達則入力Θrch 及び適応則入力Θadp を加算することで、今回の制御サイクルにおけるＳＬＤ開度補正量i/sld を求める（ＳＴＥＰ１７−１３）。そして、図１５のルーチン処理に復帰する。
【０３０２】
図１５の説明に戻って、前記ＳＴＥＰ１５−７の判断でf/dec ＝１である場合、すなわち、現在の点火時期の指令値iglog が、前記閾値IGX （図１２参照）よりも遅角側の状態となっている場合には、吸入空気量操作手段２５は、前記点火時期応動補正処理を行うべく、前記点火時期応動開度補正量θdec を前記ＳＴＥＰ１４−２で決定した開度減少単位量Δθdec づつ制御サイクル毎に増加させていく（ＳＴＥＰ１５−９）。
【０３０３】
尚、点火時期応動開度補正量θdec は、前記始動モード処理（ＳＴＥＰ３−２）で「０」に初期化されるものである。
【０３０４】
また、このＳＴＥＰ１５−９の処理を行うときは、ＳＬＤ開度補正量i/sld を求める処理は行われず、該ＳＬＤ開度補正量i/sld の値は、前記点火時期判別フラグf/dec が「１」に設定される前の値に保持される。
【０３０５】
次いで、吸入空気量操作手段２５は、ＳＴＥＰ１５−５で求めた標準開度指令値θ0 に現在のＳＬＤ開度補正量i/sld の値を加算し、さらに現在の点火時期応動開度補正量θdec を減算することで、前記吸気量Ｆ／Ｂ制御補正処理及び点火時期応動補正処理とに基づく前記予備開度指令θi/fire（＝θ0 ＋i/sld −θdec ）を算出する（ＳＴＥＰ１５−１０）。
【０３０６】
次いで、吸入空気量操作手段２５は、現在の前記学習演算終了フラグf/flrnend の値を判断する（ＳＴＥＰ１５−１１）。この学習演算終了フラグf/flrnend は、ＦＩＲＥモードの動作中（ＦＩＲＥ実行可否フラグf/fireonが「１」に設定されている状態）は、中断動作が開始されるとき（ＦＩＲＥ中断フラグf/fpauseが「１」に設定されたとき）、あるいは、点火時期の指令値iglog が前記閾値IGX （図１２参照）よりも遅角側の状態となって、前記点火時期応動補正処理を開始するとき（点火時期判別フラグf/dec が「１」に設定されたとき）に、前記基本学習補正係数vpskisldの算出処理を終了すべく「１」に設定される（前記図５のＳＴＥＰ５−１７、及び後述の図２２のＳＴＥＰ２２−８を参照）。
【０３０７】
そして、f/flrnend ＝０である場合、すなわち、前記基本学習補正係数vpskisldの算出を行うべき状態である場合には、前記学習補正処理に関して説明した通り、基本学習補正係数vpskisldを算出する（ＳＴＥＰ１５−１２）。
【０３０８】
すなわち、制御サイクル毎に、ＳＴＥＰ５−１８で求められる現在のＳＬＤ開度補正量i/sld と、内燃機関１の現在の回転数Ｎｅと、前記所定値Ｇa2とから、前記式（３４）により前記ＳＬＤ吸気補正量gair/sldを求め、それを式（３５）により累積加算することで、前記ＳＬＤ積算吸気補正量qair/sldを求める。そして、このＳＬＤ積算吸気補正量qair/sldと、前記ＳＴＥＰ１６−１０（図１６参照）で求めた制御サイクル毎の目標積算吸入空気量qair/cmdとから前記式（３６）の演算を行うことで、基本学習補正係数vpskisldを算出する。
【０３０９】
尚、この場合において、ＦＩＲＥ経過時間t/fireが前記ＳＬＤ補正制限時間TISLDLMTに達するまでは、前述の如くqair/cmd＝０とする（このときＳＬＤ積算吸気補正量qair/sldも「０」となる）ので、基本学習補正係数vpskisldの値を強制的に「１」に設定する。
【０３１０】
また、ＳＴＥＰ１５−１１でf/flrnend ＝１である場合、すなわち、前記基本学習補正係数vpskisldの算出を終了すべき状態である場合には、ＳＴＥＰ１５−１２の処理は省略され、基本学習補正係数vpskisldの算出処理は行われない（この場合、f/flrnend ＝１となる前の制御サイクルで求められた基本学習補正係数vpskisldの値がその最終値として確定される）。
【０３１１】
このようにしてＳＴＥＰ５−１１，５−１２の処理行った後、吸入空気量操作手段２５は、ＳＴＥＰ１５−１０で求めた予備開度指令θi/fireの値を所定の上限値及び下限値の間の値に制限する（θi/fire＞上限値、あるいはθi/fire＜下限値のとき、それぞれθi/fireを強制的に上限値、下限値に制限する）リミット処理を行った後（ＳＴＥＰ１５−１３）、図１４のルーチン処理に復帰する。
【０３１２】
図１４の説明に戻って、上記のように予備開度指令θi/fireを求めた後、吸入空気量操作手段２５は、この予備開度指令θi/fire（＝θ0 ＋i/sld −θdec ）に、ＳＴＥＰ１４−２で決定した前記大気圧補正係数kpa と、大気温度補正係数kta と、前回のＦＩＲＥモードの動作の終了時に決定された学習補正係数kilearn （この学習補正係数kilearn の算出については後に説明する）とを乗算することで、今回の制御サイクルにおけるバイパス開度の指令値θCMD を決定する（ＳＴＥＰ１４−４）。そして、前記図３のメインルーチン処理に復帰する。
【０３１３】
以上説明した処理が、ＦＩＲＥモードにおいて、バイパス開度の指令値θCMD を制御サイクル毎に決定するための処理である。
【０３１４】
一方、前記ＳＴＥＰ１４−１の判断で、現在のＦＩＲＥ実行可否フラグf/fireonの値が「０」である場合、すなわち、ＦＩＲＥモードの動作を行わない状態となったとき（これは、基本的には、内燃機関１の始動後、ＦＩＲＥ経過時間t/fireが、ＦＩＲＥモード制限時間TFIRELMTに達してＦＩＲＥモードの動作を終了した後の状態である）には、吸入空気量操作手段２５は、ＦＩＲＥ実行可否フラグf/fireonの前回の制御サイクルにおける値を判断する（ＳＴＥＰ１４−５）。
【０３１５】
このとき、ＦＩＲＥ実行可否フラグf/fireonの前回の制御サイクルにおける値が「１」である場合、すなわち、前回の制御サイクルまでＦＩＲＥモードの動作を行っていた場合（ＦＩＲＥモードを終了した直後の状態）には、直前まで行っていたＦＩＲＥモードにおいて前記ＳＴＥＰ１５−１２の処理により最終的に求められた前記基本学習補正係数vpskisldから、図１８のフローチャートに示すように次回のＦＩＲＥモードの動作に際して開度指令を補正する（ＳＴＥＰ１４−４を参照）ための学習補正係数kilearn を決定（更新）する（ＳＴＥＰ１４−６）。
【０３１６】
すなわち、図１８を参照して、まず、前述した基本学習補正係数vpskisldの算出を最終的に終了した時のＦＩＲＥ経過時間t/fireを表す前記学習終了時刻パラメータt/kil が、あらかじめ定めた所定値TMKILLMT以上であるか否かを判断する（ＳＴＥＰ１８−１）。ここで、学習終了時刻パラメータt/kil は、前記図５に示した如く、前記学習演算終了フラグf/flrnend が「０」から「１」に切り換えられた際のＦＩＲＥ経過時間t/fireである。つまり、ＦＩＲＥモードの中断動作が行われたときは、該中断動作の開始時までのＦＩＲＥ経過時間t/fireであ、ＦＩＲＥモードの動作中に前記点火時期応動補正処理を行うときは、その処理の開始時までのＦＩＲＥ経過時間t/fireである。そして、ＦＩＲＥモードの中断動作や点火時期応動補正処理が行われることなくＦＩＲＥモードの動作が終了したときには、その終了時のＦＩＲＥ経過時間t/fire（これは通常的にはＦＩＲＥモード制限時間TFIRELMTである）である。
【０３１７】
別の言い方をすれば、学習終了時刻パラメータt/kil は、前記ＳＬＤ開度補正量i/sld の算出処理及びこれに応じた標準開度指令値θ0 の補正が連続して行われたＦＩＲＥ経過時間t/fireである。
【０３１８】
そして、ＳＴＥＰ１８−１判断で、t/kil ＜TMKILLMTである場合には、最終的に得られた基本学習補正係数vpskisldの信頼性が乏しいと考えられるので、学習補正係数kilearn を現状の値に維持したまま図１４の処理に復帰する。
【０３１９】
一方、t/kil ≧TMKILLMTである場合、すなわち、ＳＬＤ開度補正量i/sld の算出処理及びこれに応じた標準開度指令値θ0 の補正がある程度長い時間、連続して行われた場合には、前回の制御サイクルまで行われていたＦＩＲＥモードの動作において最終的に得られた基本学習補正係数vpskisldから前記式（３７）のなまし演算処理（フィルタリング処理）によって新たな学習補正係数kilearn(j)を求める（ＳＴＥＰ１８−２）。そして、この学習補正係数kilearn(j)を所定の上限値及び下限値の間の値に制限する（kilearn(j)＞上限値、あるいはkilearn(j)＜下限値のとき、それぞれkilearn(j)を強制的に上限値、下限値に制限する）リミット処理を行った後（ＳＴＥＰ１８−３）、図１４のルーチン処理に復帰する。
【０３２０】
尚、学習補正係数kilearn(j)の初期値は「１」である。また、この学習補正係数kilearn(j)の値は、本実施形態のシステムの運転を停止しても失われることのないようにＥＥＰＲＯＭ等の不揮発性メモリに記憶保持しておく。
【０３２１】
図１４の説明に戻って、ＳＴＥＰ１４−５でＦＩＲＥ実行可否フラグf/fireonの前回の制御サイクルにおける値が「１」である場合には、前述の如く学習補正係数kilearn が更新されるが、ＦＩＲＥ実行可否フラグf/fireonの前回の制御サイクルにおける値が「０」である場合は、既に学習補正係数kilearn が更新されているので、ＳＴＥＰ１４−６の処理は省略される。すなわち、学習補正係数kilearn の更新処理は、ＦＩＲＥモードが終了した直後の制御サイクルにおいてだけ行われ、その更新された学習補正係数kilearn が次回のＦＩＲＥモードの動作に際して開度指令を補正するために使用される。
【０３２２】
このようにしてＳＴＥＰ１４−５、１４−６の処理を行った後、吸入空気量操作手段２５は、開度指令θCMD を通常モード用の開度指令に設定する（ＳＴＥＰ１４−７）。この開度指令は、基本的には、ＦＩＲＥモードにおける開度指令θCMD よりも小さく、内燃機関１の通常的な運転を行うべく所要の値に設定される。
【０３２３】
以上、図１４〜図１８を参照して説明した処理が、前記図４のＳＴＥＰ４−５で開度指令（バイパス開度の指令値）θCMD を制御サイクル毎に生成する処理の詳細であり、このようにして決定された開度指令θCMD は前記バイパス弁アクチュエータ２４に与えられる。そして、該バイパス弁アクチュエータ２４は、与えられた開度指令θCMD に従ってバイパス弁７の開度を操作する。
【０３２４】
次に、前記図４のＳＴＥＰ４−６における点火時期の指令値iglog の生成処理について説明する。
【０３２５】
ここで、この処理の具体的な内容を説明する前に、この処理の基本的な内容を説明しておく。
【０３２６】
本実施形態のシステムでは、ＦＩＲＥモードにおける前述のような吸入空気量の増量制御によって内燃機関１の回転数Ｎｅ（実回転数）が上昇傾向となるので、該回転数Ｎｅを前記点火時期操作回転数Ｆ／Ｂ制御によって所要の目標回転数ne/fire にフィードバック制御し、該点火時期を遅角側に補正する（点火時期操作回転数Ｆ／Ｂ制御）。この場合、後述する如く、回転数Ｎｅと目標回転数ne/fire との偏差の減衰速度（これはフィードバック制御のゲインに相当する）を、適宜可変的に設定することが好ましい。
【０３２７】
このために、本実施形態のシステムにおける点火時期操作回転数Ｆ／Ｂ制御にあっては、前述の吸気量Ｆ／Ｂ制御補正処理の場合と同様に、上記偏差の減衰速度を指定可能な応答指定型制御であるスライディングモード制御（本実施形態では適応スライディングモード制御）を用いる。
【０３２８】
この適応スライディングモード制御（以下、点火時期側適応ＳＬＤ制御という）を用いた点火時期操作回転数Ｆ／Ｂ制御の処理、すなわち、回転数Ｎｅを目標回転数ne/fire に収束させるように点火時期の指令値iglog を生成する処理のアルゴリズムは次のように構築されている。
【０３２９】
まず、本実施形態では、点火時期側適応ＳＬＤ制御の制御対象を、点火時期の指令値iglog を表すデータから、回転数Ｎｅ（実回転数）を表すデータを生成する系と考え、それを離散系（離散時間系）でモデル化しておく。
【０３３０】
この場合、本実施形態では、点火時期の指令値iglog と所定の基準指令値ig0 と偏差DIG （＝iglog −ig0 。以下、点火時期偏差指令値DIG という）を点火時期の指令値iglog を表すデータとして用いると共に、回転数Ｎｅと所定の基準回転数Ｎe0との偏差DNE （＝Ｎｅ−Ｎe0。以下、偏差回転数DNE という）を回転数Ｎｅを表すデータとして用いる。
【０３３１】
尚、本実施形態では、点火時期の基準指令値ig0 として、前記図３で示した基本指令値igbase（内燃機関１のＦＩＲＥモード以外の通常的な運転時における点火時期の指令値）を用いる（ig0 ＝igbase）。従って、前記図３で示した補正量DIG が上記点火時期偏差指令値DIG である。また、本実施形態では、回転数Ｎｅに係わる基準回転数Ｎe0として、前記図３で示した所定のアイドリング回転数NOBJを用いる（Ｎe0＝NOBJ）。
【０３３２】
そして、本実施形態で、上記の点火時期偏差指令値DIG と、偏差回転数DNE とを用いて、点火時期側適応ＳＬＤ制御の制御対象のモデルを次式（３８）のように離散系モデル（本実施形態では２次の自己回帰モデル）で表現しておく。
【０３３３】
【数３８】

【０３３４】
この式（３８）により表現した制御対象のモデル（以下、回転数制御対象モデルという）は、該回転数制御対象モデルの出力としての各制御サイクルにおける偏差回転数DNE(k+1)を、それより過去の偏差回転数DNE の時系列データDNE(K)，DNE(k-1)、並びに該回転数制御対象モデルの入力としての点火時期偏差指令値DIG(k)を用いて表現したものである。
【０３３５】
また、式（３８）で、偏差回転数DNE(K)，DNE(k-1)にそれぞれ係る係数ｃ1 ，ｃ2 と、点火時期偏差指令値DIG(k)に係る係数ｄ1 とは、回転数制御対象モデルの実際の挙動特性を規定するモデルパラメータであり、これらのモデルパラメータｃ1 ，ｃ2 ，ｄ1 は、回転数制御対象モデルの挙動特性と、該モデルにより表現した実際の制御対象の挙動特性とが整合するようにあらかじめ実験やシミュレーション等を通じて同定しておく。
【０３３６】
この場合、回転数制御対象モデルは離散系モデルであるため、種々の公知の同定アルゴリズム（例えば回転数制御対象モデル上で生成される偏差回転数DNE(k+1)と実際の偏差回転数との間の誤差が最小になるように最小二乗法によりモデルパラメータｃ1 ，ｃ2 ，ｄ1 を同定するアルゴリズム等）を用いてモデルパラメータｃ1 ，ｃ2 ，ｄ1 を比較的容易に同定することが可能である。
【０３３７】
このようにして定めた回転数制御対象モデルに基づいて、点火時期側適応ＳＬＤ制御の処理のアルゴリズムは次のように構築する。
【０３３８】
すなわち、点火時期側適応ＳＬＤ制御では、スライディングモード制御に必要な切換関数σ2 を、前述の吸気側適応スライディングモード制御の場合と同様に、前記偏差回転数DNE とその目標値dne との偏差Ｅｎ＝DNE −dne （以下、回転数偏差Ｅｎという）の制御サイクル毎の時系列データＥn(k)，Ｅn(k-1)を変数とする次式（３９）の線形関数により定義する。尚、偏差回転数DNE の目標値dne （以下、偏差目標回転数dne という）は、前記図３で示した目標回転数ne/fire と前記基準回転数Ｎe0（＝アイドリング回転数NOBJ）との偏差（＝ne/fire −Ｎe0）である。従って、上記回転数偏差Ｅｎ＝DNE −dne は、回転数Ｎｅ（実回転数）と目標回転数ne/fire との偏差（＝Ｎｅ−ne/fire ）と同じである。
【０３３９】
【数３９】

【０３４０】
ここで、式（３９）中のｓ3 ，ｓ4 は切換関数σ2 の各項の係数パラメータであり、次の条件を満たすように設定する。
【０３４１】
【数４０】

【０３４２】
尚、本実施形態では、簡略化のためにｓ3 ＝１としている。また、係数パラメータｓ4 の値（より一般的には、ｓ4 ／ｓ3 ＝pole/ig の値）を可変的に設定するのであるが、これについては後述する。
【０３４３】
このように切換関数σ2 を定義したとき、吸気側適応スライディングモード制御の場合と同様に、前記回転数偏差Ｅｎの時系列データＥn(k)，Ｅn(k-1)の組から成る状態量（Ｅn(k)，Ｅn(k-1)）を、σ2 ＝０なる関係式によって定義される切換線上に収束させ、その収束状態を維持すると、状態量（Ｅn(k)，Ｅn(k-1)）を、外乱等の影響によらずに極めて安定に切換線σ2 ＝０上の平衡点、すなわち、Ｅn(k)＝Ｅn(k-1)＝０となる点に収束させることができる。
【０３４４】
また、前記偏差回転数DNE を前記偏差目標回転数dne に収束させる（回転数Ｎｅを目標回転数ne/fire に収束させる）ために、前記式（３９）によりモデル化した制御対象に与えるべき入力として点火時期側適応ＳＬＤ制御が生成する制御入力、すなわち点火時期偏差指令値DIG は、吸気側適応スライディングモード制御の場合と同様に、等価制御入力DIGeq と到達則入力DIGrchと適応則入力DIGadpとの総和である（次式（４１）を参照）。
【０３４５】
【数４１】

【０３４６】
そして、これらの等価制御入力DIGeq 、到達則入力DIGrch及び適応則入力DIGadpは、吸気側適応スライディングモード制御の場合と同様に、それぞれ次式（４２）〜（４４）により与えられる。
【０３４７】
【数４２】

【０３４８】
【数４３】

【０３４９】
【数４４】

【０３５０】
そして、この場合、式（４３）中の係数Ｆ3 （到達則のゲインを規定する係数）は、次式（４５）、より好ましくは（４５）’の条件を満たすようにあらかじめ設定しておく。
【０３５１】
また、式（４４）中の係数Ｆ4 （適応則のゲインを規定する係数）は、次式（４６）の条件を満たすようにあらかじめ設定しておく。尚、式（４６）中のΔＴは制御サイクル（制御周期）である。
【０３５２】
【数４５】

【０３５３】
【数４６】

【０３５４】
本実施形態における点火時期側適応ＳＬＤ制御では、制御サイクル毎に、前記式（４２）〜（４４）により等価制御入力DIGeq 、到達則入力DIGrch及び適応則入力DIGadpをそれぞれ求め、それらの総和を演算する（式（４１））ことで、点火時期偏差指令値DIG を求める。そして、この点火時期偏差指令値DIG を、次式（４７）のように前記基準指令値ig0 、すなわち内燃機関１の通常的な運転時の基本指令値igbaseに加算することで、点火時期の指令値iglog を決定する。
【０３５５】
【数４７】

【０３５６】
この場合、前記等価制御入力DIGeq や切換関数σ2 の値を求めるために必要な偏差目標回転数dne は次のように求める。
【０３５７】
すなわち、本実施形態では、内燃機関１の回転数Ｎｅ（実回転数）が、前記図３に示した設定回転数（＝NOBJ＋NEFSLDS ）に達した時、又は、前記ＦＩＲＥ経過時間t/fireがあらかじめ定めた所定値TSLDIGSTに達した時から点火時期操作回転数Ｆ／Ｂ制御を開始する。そして、ＦＩＲＥモードにおける内燃機関１の目標回転数ne/fire は、点火時期操作回転数Ｆ／Ｂ制御の開始時からの経過時間Δt/nfb （以下、回転数Ｆ／Ｂ経過時間Δt/nfb という）に応じて次式（４８）により設定する。
【０３５８】
【数４８】

【０３５９】
ここで、式（４８）中のK/NEは、目標回転数ne/fire の時間的減少度合い（傾き）を規定する一定の所定値（＞０）である。
【０３６０】
尚、この場合において、式（４８）の右辺の演算結果がアイドリング回転数NOBJを下回る場合（Δt/nfb ＞NEFSLDS ／K/NEの場合）には、目標回転数ne/fire をアイドリング回転数NOBJに固定する。
【０３６１】
これにより、目標回転数ne/fire は、点火時期操作回転数Ｆ／Ｂ制御の開始後、前記設定回転数（＝NOBJ＋NEFSLDS ）からアイドリング回転数NOBJに向かって直線的に徐々に減少し、該アイドリング回転数NOBJに達した後は、該アイドリング回転数NOBJに維持される。
【０３６２】
従って、本実施形態では、各制御サイクルにおける目標回転数ne/fire(k)は、該制御サイクルにおける前記回転数Ｆ／Ｂ経過時間Δt/nfb から式（４８）により求められ、この目標回転数ne/fire(k)からアイドリング回転数NOBJを減算することで、各制御サイクルにおける偏差目標回転数dne(k)（＝ne/fire(k)−NOBJ）が求められる。そして、この偏差目標回転数dne(k)と、その前回値dne(k-1)（＝ne/fire(k-1)−NOBJ）、すなわち前回の制御サイクルで求めた偏差目標回転数dne(k-1)とを用いることで、制御サイクル毎に前記式（３９）により切換関数σ2(k)の値を求めることができる。そして、この切換関数σ2 の値を用いることで、前記式（４３）、（４４）により到達則入力DIGrch及び適応則入力DIGadpを求めることができる。
【０３６３】
また、本実施形態では、制御サイクル（ＴＤＣ）は、内燃機関１の回転数Ｎｅに反比例するので、現在の回転数Ｎｅから１制御サイクル分の時間ΔＴ（∝１／Ｎｅ）を求め、該時間ΔＴを現在の回転数Ｆ／Ｂ経過時間Δt/nfb に加算することで、次回の制御サイクルにおける回転数Ｆ／Ｂ経過時間（＝Δt/nfb ＋ΔＴ）を予測することができる（以下、この予測値を回転数Ｆ／Ｂ予測経過時間Δt/nfbpreという）。そして、この回転数Ｆ／Ｂ予測経過時間Δt/nfbpreを前記式（４８）に適用する（式（４８）の右辺のΔt/nfb にΔt/nfbpreを代入する）ことで、次式（４９）により次回の制御サイクルにおける目標回転数ne/fire(k+1)を求めることができる。
【０３６４】
【数４９】

【０３６５】
さらに、該目標回転数ne/fire(k+1)からアイドリング回転数NOBJを減算することで、次回の制御サイクルにおける偏差目標回転数dne(k+1)（＝ne/fire(k+1)−NOBJ）が求められる。
【０３６６】
そして、このようにして求められる偏差目標回転数dne(k+1)と、前述の如く求められる偏差目標回転数dne(k)，dne(k-1)とを用いて前記式（４２）の演算を行うことで、等価制御入力DIGeq を求めることができる。
【０３６７】
尚、本実施形態では、ＦＩＲＥモードの中断動作中は、点火時期操作回転数Ｆ／Ｂ制御を中断し、点火時期の指令値iglog を内燃機関１の通常的な運転時の基本指令値igbaseに戻す。この場合、中断動作が開始してから、式（４７）の点火時期偏差指令値DIG の値（絶対値）を、その値が「０」になるまで制御サイクル毎に所定の単位値dec/igづつ減少させていく（徐々に「０」に近づけていく）ことで、点火時期の指令値iglog を徐々に基本指令値igbaseに戻すようにしている。そして、中断動作が解除されたときには、直ちに点火時期操作回転数Ｆ／Ｂ制御を再開するようにしている。
【０３６８】
また、本実施形態では、ＦＩＲＥモードの動作が終了し（ＦＩＲＥ実行可否フラグf/fireonが「１」から「０」になる）、システムの動作モードが通常モードに移行する際にも、点火時期の指令値iglog を徐々に基本指令値igbaseに戻すようにしている。
【０３６９】
以上説明した内容が、本実施形態における点火時期操作回転数Ｆ／Ｂ制御の処理の基本的内容である。
【０３７０】
ところで、応答指定型制御である前記点火時期側適応ＳＬＤ制御にあっては、前記吸気側適応ＳＬＤ制御と同様、前記切換関数σ2 の係数パラメータｓ3 ，ｓ4 の比（ｓ4 ／ｓ3 ）の値（以下、この比の値をポールpole/ig という）によって、前記回転数偏差Ｅｎ（＝DNE −dne ＝Ｎｅ−ne/fire ）の減衰速度を指定することができる。すなわち、ポールpole/ig （＝ｓ4 ／ｓ3 ）の絶対値を「１」よりも小さい範囲で「０」に近づけていく程、回転数偏差Ｅｎの減衰速度は速くなる。尚、本実施形態では、前記吸気側適応ＳＬＤ制御の場合と同様、回転数偏差Ｅｎの振動的な減衰を避けるために、ｓ4 ／ｓ3 ＝pole/ig ＜０としている。
【０３７１】
このような点火時期側適応ＳＬＤ制御の応答指定特性を利用し、本実施形態では次のように回転数偏差Ｅｎの減衰速度を可変的に設定する。
【０３７２】
すなわち、点火時期の変化に対する内燃機関１の回転数Ｎｅの変化は該点火時期が遅角側である程、大きくなる傾向がある。このため、回転数Ｎｅの目標回転数ne/fire への制御を安定して行うためには、回転数偏差Ｅｎの減衰速度を、操作している点火時期が遅角側である程、遅くしてやる（ポールpole/ig の絶対値を大きくする）ことが好ましいと考えられる。
【０３７３】
そこで、本実施形態では、制御サイクル毎に、現在の点火時期の指令値iglog （前回の制御サイクルで決定された指令値iglog ）から、図１９に示す如くあらかじめ定めたデータテーブルに基づいてポールpole/ig の基本値pole/igtblを求め、基本的には、この基本値pole/igtblをポールpole/ig の値として設定する。この場合、基本値pole/igtbl（以下、点火時期対応ポール基本値pole/igtblという）は、点火時期の指令値iglog が遅角側である程、その絶対値｜pole/igtbl｜を大きくするように設定されている（但し、−１＜pole/igtbl＜０）。
【０３７４】
尚、点火時期対応ポール基本値pole/igtblは、基本的には、前記吸気側適応ＳＬＤ制御に関して前記図９のデータテーブルにより決定するポールテーブル値pole/itbl よりも「０」側に近い値に設定されている。換言すれば、点火時期側適応ＳＬＤ制御に関して点火時期対応ポール基本値pole/igtblにより規定される前記回転数偏差Ｅｎの減衰速度は、前記吸気側適応ＳＬＤ制御に関してポールテーブル値pole/itbl に規定される前記吸気偏差Ｅq の減衰速度よりも速くなるように点火時期対応ポール基本値pole/igtblが設定されている。
【０３７５】
また、本実施形態では、点火時期操作回転数Ｆ／Ｂ制御を開始する際に、前述した点火時期側適応ＳＬＤ制御の処理によって回転数Ｎｅを目標回転数ne/fire に収束させるべく点火時期を急激に変化させると、内燃機関１の燃焼状態が悪化する虞れがある。従って、点火時期操作回転数Ｆ／Ｂ制御の初期段階では、回転数偏差Ｅｎの減衰速度を遅めにするようにポールpole/ig の値を設定することが好ましい。
【０３７６】
そこで、本実施形態では、制御サイクル毎に、前記回転数Ｆ／Ｂ経過時間Δt/nfb （点火時期操作回転数Ｆ／Ｂ制御を開始してからの経過時間）から、図２０に示す如くあらかじめ定めたデータテーブル（タイムテーブル）に基づいて点火時期対応ポール基本値pole/igtblを回転数Ｆ／Ｂ経過時間Δt/nfb に応じて補正（乗算補正）するための補正係数kigtを求め、この補正係数kigt（以下、時間対応補正係数kigtという）を点火時期対応ポール基本値pole/igtblに乗算することで、該点火時期対応ポール基本値pole/igtblを補正するようにしている。
【０３７７】
この場合、図２１のデータテーブルでは、時間対応補正係数kigtは、点火時期操作回転数Ｆ／Ｂ制御の初期段階（回転数Ｆ／Ｂ経過時間Δt/nfb が所定値T/NFBXに達するまで）では、点火時期対応ポール基本値pole/igtblの絶対値を若干大きくする方向（回転数偏差Ｅｎの減衰速度を遅くする方向）に該点火時期対応ポール基本値pole/igtblを補正するような値（＞１）に定められている。しかもこのとき、時間対応補正係数kigtは、回転数Ｆ／Ｂ経過時間Δt/nfb が短い程、大きな値に定められ、これにより、回転数Ｆ／Ｂ経過時間Δt/nfb が短い程、点火時期対応ポール基本値pole/igtblの絶対値がより大きくなる（回転数偏差Ｅｎの減衰速度がより遅くなる）ように点火時期対応ポール基本値pole/igtblを補正するようにしている。そして、回転数Ｆ／Ｂ経過時間Δt/nfb が所定値T/NFBXに達した後は、時間対応補正係数kigtは「１」に保持され、この状態では、点火時期対応ポール基本値pole/igtblを補正しないようにしている。
【０３７８】
さらに、本実施形態では、ＦＩＲＥモードの中断動作中は、点火時期操作回転数Ｆ／Ｂ制御は中断され、回転数Ｎｅのフィードバック制御を行わないため、該中断動作が解除され、点火時期操作回転数Ｆ／Ｂ制御を再開する際に、内燃機関１の回転数が目標回転数ne/fire よりも大幅に高いものとなっていることがある。そして、このような状態では、前記回転数偏差Ｅｎが大きなものとなるため、点火時期操作回転数Ｆ／Ｂ制御によって前述の如く求められる点火時期の指令値iglog が急激に遅角側に過大なものとなり、内燃機関１の燃焼状態が悪化する虞れがある。従って、ＦＩＲＥモードの中断動作の解除によって点火時期操作回転数Ｆ／Ｂ制御を再開する際には、内燃機関１の回転数Ｎｅが目標回転数ne/fire に対して大きく高回転側に離間しているような状況では、回転数偏差Ｅｎの減衰速度を遅くし、ひいては、点火時期の指令値iglog を過大に遅角側にするような前記点火時期偏差指令値DIG が求められるのを回避することが好ましいと考えられる。
【０３７９】
そこで、本実施形態では、制御サイクル毎に、内燃機関１の現在の回転数Ｎｅから、図２１に示す如くあらかじめ定めたデータテーブルに基づいて、点火時期対応ポール基本値pole/igtblを回転数Ｎｅに応じて補正（乗算補正）するための補正係数kigne （以下、回転数対応補正係数kigne という）を求める。この場合、図２１のデータテーブルでは、回転数対応補正係数kigne は基本的には、回転数Ｎｅが高い程（回転数Ｎｅが目標回転数ne/fire に対して高回転側に離間している程）、点火時期対応ポール基本値pole/igtblの絶対値を大きくする方向（回転数偏差Ｅｎの減衰速度を遅くする方向）に該点火時期対応ポール基本値pole/igtblを補正するような値（＞１）に設定されている。そして、本実施形態では、回転数対応補正係数kigne に基づく点火時期対応ポール基本値pole/igtblの補正を、点火時期操作回転数Ｆ／Ｂ制御の再開後、所定期間XCNT（これは、前記図３のＳＴＥＰ５−１８で設定するカウントダウンタイマcnt/igvpl の初期値である）だけ行うために、次式（５０）により回転数対応補正係数kigne を修正し、この修正した補正係数kignef（以下、回転数対応修正補正係数kignefという）を点火時期対応ポール基本値pole/igtblに乗算することで、該点火時期対応ポール基本値pole/igtblを補正する。
【０３８０】
【数５０】

【０３８１】
ここで、式（５０）中のカウントダウンタイマcnt/igvpl は、図３のＳＴＥＰ５−１８によって、中断動作中（ＦＩＲＥ中断フラグf/fpause＝１の状態）に常時、上記所定期間XCNTが設定される。そして、中断動作の解除により点火時期操作回転数Ｆ／Ｂ制御が再開されると、カウントダウンタイマcnt/igvpl の値は、所定期間XCNTの値から、制御サイクル毎に所定値づつ減少し、最終的に「０」になった後はその値に保持される。従って、式（５０）により求まる回転数対応修正補正係数kignefは、点火時期操作回転数Ｆ／Ｂ制御が再開してから所定期間XCNTが経過するまでは、回転数Ｎｅに応じた値（≧１）であるが、該所定期間XCNTが経過した後は、kignef＝１である（このとき、回転数対応修正補正係数kignefによる点火時期対応ポール基本値pole/igtblの補正は行われなくなる）。
【０３８２】
尚、カウントダウンタイマcnt/igvpl は、前記始動モード処理（ＳＴＥＰ３−２）において「０」に初期化されるもので、ＦＩＲＥモードの中断動作中と、該中断動作が終了してから所定期間XCNTが経過するまでの期間とを除いて値が「０」に保持される。
【０３８３】
本実施形態では、制御サイクル毎に、前述のように求めた点火時期対応ポール基本値pole/igtblに、次式（５１）の如く前記時間対応補正係数kigt及び回転数対応修正補正係数kignefを乗算してなる値を、最終的なポールpole/ig の値として設定する。
【０３８４】
【数５１】

【０３８５】
尚、時間対応補正係数kigtは、点火時期操作回転数Ｆ／Ｂ制御の開始直後の初期段階においてのみkigt＞１となり、これ以外の状態では、kigt＝１である。また、回転数対応修正補正係数kignefは、ＦＩＲＥモードの中断動作の解除直後の初期段階で、しかも回転数Ｎｅが比較的高い場合のみkignef＞１となり、これ以外の状態ではkignef＝１である。従って、式（５１）により設定されるポールpole/ig の値は、通常的には、点火時期対応ポール基本値pole/igtblである。
【０３８６】
以上説明した内容をふまえて、前記図４のＳＴＥＰ４−６において前記点火時期操作手段２６が行う点火時期の指令値iglog の生成処理を次に具体的に説明する。
【０３８７】
図２２のフローチャートを参照して、前記ＳＴＥＰ４−６では、点火時期操作手段２６は、まず、点火時期の基本指令値igbaseを決定する（ＳＴＥＰ２２−１）。この場合、該基本指令値igbaseは、例えば内燃機関１の現在の回転数Ｎｅや吸気圧Ｐｂ、機関温度Ｔｗ、大気温度Ｔａ等から、あらかじめ定めたマップや演算式を用いて求める。
【０３８８】
次いで、点火時期操作手段２６は、前記点火時期偏差指令値DIG を決定するための処理を図２３のフローチャートに示すように実行する（ＳＴＥＰ２２−２）。
【０３８９】
すなわち、ＳＴＥＰ２２−２の処理では、まず、現在のＦＩＲＥ実行可否フラグf/fireonの値を判断する（ＳＴＥＰ２３−１）。
【０３９０】
このとき、f/fireon＝１、すなわち、ＦＩＲＥモードの動作を行うべき状態である場合には、点火時期操作手段２６は点火時期操作回転数Ｆ／Ｂ制御を行うか否かをそれぞれ値「１」、「０」で表すフラグf/nefb（以下、回転数Ｆ／Ｂ実行可否フラグf/nefbという）の値を判断する（ＳＴＥＰ２３−２）。
【０３９１】
この回転数Ｆ／Ｂ実行可否フラグf/nefbは、前記始動モード処理（ＳＴＥＰ４−２）において「０」に初期化されるものである。
【０３９２】
このＳＴＥＰ２３−２において、f/nefb＝０、すなわち、点火時期操作回転数Ｆ／Ｂ制御をまだ行うべき状態でない場合には、前記回転数Ｆ／Ｂ経過時間Δt/nfb の値を「０」に初期化する（ＳＴＥＰ２３−３）。該回転数Ｆ／Ｂ経過時間Δt/nfb は、ＳＴＥＰ２３−２でf/nefb＝１となって、点火時期操作回転数Ｆ／Ｂ制御を行うべき状態となった制御サイクルから計時を開始するものであり、ＳＴＥＰ２３−２でf/nefb＝１である場合には、ＳＴＥＰ２３−３の処理は省略される。
【０３９３】
次いで、点火時期操作手段２６は、今回の制御サイクルにおける目標回転数ne/fire(k)と、次回の制御サイクルにおける目標回転数ne/fire(k+1)とを求める（ＳＴＥＰ２３−４）。この場合、今回の制御サイクルにおける目標回転数ne/fire(k)は、現在の回転数Ｆ／Ｂ経過時間Δt/nfb から前記式（４８）により求められる。また、次回の制御サイクルにおける目標回転数ne/fire(k+1)に関しては、現在の内燃機関１の回転数Ｎｅ（回転数センサ１４の検出値）から把握される１制御サイクルの時間ΔＴを現在の回転数Ｆ／Ｂ経過時間Δt/nfb に加算することで回転数Ｆ／Ｂ予測経過時間Δt/nfbpreを求め、この回転数Ｆ／Ｂ予測経過時間Δt/nfbpreから前記式（４９）により目標回転数ne/fire(k+1)が求められる。
【０３９４】
尚、今回の制御サイクルにおける目標回転数ne/fire(k)は、回転数Ｆ／Ｂ経過時間Δt/nfb がＳＴＥＰ２３−３で「０」に設定されている間（f/nefb＝０の状態）は、前記設定回転数（NOBJ＋NEFSLDS ）である。
【０３９５】
次いで、点火時期操作手段２６は、内燃機関１の現在の回転数Ｎｅが現在の目標回転数ne/fire(k)以上となったか否か（ＳＴＥＰ２３−５）、現在のＦＩＲＥ経過時間t/fireが所定値TSLDIGST（一定値）以上となったか否か（ＳＴＥＰ２３−６）を順次判断する。
【０３９６】
そして、ＳＴＥＰ２３−５、２３−６のいずれかの条件が成立したときには、回転数Ｆ／Ｂ実行可否フラグf/nefbの値を「１」に設定する（ＳＴＥＰ２３−７）。これにより、ＦＩＲＥモードの動作が開始してから、内燃機関１の回転数Ｎｅが前記設定回転数（NOBJ＋NEFSLDS ）に達した時、又は、ＦＩＲＥ経過時間t/fireが所定値TSLDIGSTに達した時に回転数Ｆ／Ｂ実行可否フラグf/nefbが「１」に設定され、点火時期操作回転数Ｆ／Ｂ制御を行い得る状態となる。
【０３９７】
尚、ＳＴＥＰ２３−５、２３−６のいずれかの条件も成立しない状態では、ＳＴＥＰ２３−７の処理は省略され、回転数Ｆ／Ｂ実行可否フラグf/nefbの値は「０」に維持される。また、回転数Ｆ／Ｂ実行可否フラグf/nefbは、「１」に設定された後にＦＩＲＥモードの動作中に「０」に戻されることはない。
【０３９８】
次いで、点火時期操作手段２６は、内燃機関１の現在の回転数Ｎｅと、ＳＴＥＰ２３−４で求めた現在の目標回転数ne/fire(k)とから今回の制御サイクルにおける前記回転数偏差Ｅｎ(k) （＝Ｎｅ−ne/fire(k)）を算出する（ＳＴＥＰ２３−８）。
【０３９９】
次いで、点火時期操作手段２６は、前記ＦＩＲＥ中断フラグf/fpauseの値を判断し（ＳＴＥＰ２３−９）、f/fpause＝０、すなわち、ＦＩＲＥモードの中断動作を行う状態でない場合には、さらに回転数Ｆ／Ｂ実行可否フラグf/nefbの値を判断する（ＳＴＥＰ２３−１０）。
【０４００】
このとき、f/nefb＝０である場合には、今回の制御サイクルにおける点火時期偏差指令値DIG(k)の値を「０」として（ＳＴＥＰ２３−１１）、図２２のルーチン処理に復帰する。
【０４０１】
一方、ＳＴＥＰ２３−１０でf/nefb＝１である場合には、点火時期操作回転数Ｆ／Ｂ制御のための点火時期偏差指令値DIG(k)を算出する（ＳＴＥＰ２３−１２）。
【０４０２】
この点火時期偏差指令値DIG(k)の算出は次のように行われる。
【０４０３】
すなわち図２４のフローチャートを参照して、まず、点火時期の現在の指令値iglog(k-1)（前回の制御サイクルで決定された指令値）から、前記図１９のデータテーブルにより前記点火時期対応ポール基本値pole/igtblを求める（ＳＴＥＰ２４−１）。
【０４０４】
また、現在の回転数Ｆ／Ｂ経過時間Δt/nfb から、前記図２０のデータテーブルにより前記時間対応補正係数kigtを求める（ＳＴＥＰ２４−２）。
【０４０５】
また、内燃機関１の現在の回転数Ｎｅから、前記図２１のデータテーブルにより前記回転数対応補正係数kigne を求め（ＳＴＥＰ２４−３）、さらにこの回転数対応補正係数kigne と、前記カウントダウンタイマcnt/igvpl の現在の値と、ＦＩＲＥモードの中断動作の終了後のカウントダウンタイマcnt/igvpl の初期値としてあらかじめ定めた所定期間XCNTとから前記式（５０）により前記回転数対応修正補正係数kignefを求める（ＳＴＥＰ２４−４）。
【０４０６】
そして、ＳＴＥＰ２４−１で求めた点火時期対応ポール基本値pole/igtblに、ＳＴＥＰ２４−２及び２４−４でそれぞれ求めた時間対応補正係数kigt及び回転数対応修正補正係数kignefを乗算する（式（５１）の演算を行う）ことで、今回の制御サイクルにおけるポールpole/ig の値を決定する（ＳＴＥＰ２４−５）。
【０４０７】
次いで、点火時期操作手段２６は、今回及び前回の制御サイクルにおいてそれぞれ前記ＳＴＥＰ２３−８で求めた回転数偏差Ｅｎ(k) ，Ｅｎ(k-1) と、上記ＳＴＥＰ２４−５で求めたポールpole/ig の今回値とから、前記式（３９）の演算により今回の制御サイクルにおける前記切換関数σ2(k)を求める（ＳＴＥＰ２４−６）。この場合において、式（３９）中の係数パラメータｓ3 ，ｓ4 はそれぞれ、「１」、「pole/ig 」である。
【０４０８】
次いで、点火時期操作手段２６は、前記式（４２）〜（４４）により、今回の制御サイクルにおける等価制御入力DIGeq 、到達則入力DIGrch、及び適応則入力DIDadpをそれぞれ求める（ＳＴＥＰ２４−７）。
【０４０９】
より具体的には、等価制御入力DIGeq に関しては、まず、今回の制御サイクルにおいて前記ＳＴＥＰ２３−４で求めた目標回転数ne/fire(k)及びne/fire(k+1)と、前回の制御サイクルにおいてＳＴＥＰ２３−４で求めた目標回転数ne/fire(k-1)とから、それらの目標回転数ne/fire(k)，ne/fire(k+1)，ne/fire(k-1)の前記基準回転数Ｎe0（＝アイドリング回転数NOBJ）に対する偏差、すなわち前記偏差回転数dne(k)，dne(k+1)，dne(k-1)を求める。そして、この偏差回転数dne(k)，dne(k+1)，dne(k-1)と、今回及び前回の制御サイクルにおいてそれぞれ前記ＳＴＥＰ２３−８で求めた回転数偏差Ｅｎ(k) ，Ｅｎ(k-1) と、ＳＴＥＰ２４−５で求めたポールpole/ig の今回値とを用いて式（４２）の演算を行うことで、等価制御入力DIGeq(k)が求められる。この場合において、式（４２）中の係数パラメータｓ3 ，ｓ4 はそれぞれ、「１」、「pole/ig 」である。また、式（４２）中のモデルパラメータｃ1 ，ｃ2 ，ｄ1 は前記回転数制御対象モデル（式（３８）を参照）についてあらかじめ同定した所定値である。
【０４１０】
また、到達則入力DIGrchに関しては、前記ＳＴＥＰ２４−６で求めた切換関数σ2(k)の値を用いて式（４３）の演算を行うことで、到達則入力DIGrch(k) が求められる。この場合において式（４３）中の係数パラメータｓ3 は「１」であり、また、式（４３）中の係数Ｆ3 は前記式（４５）もしくは（４５）’の条件を満たすようにあらかじめ設定した所定値である。
【０４１１】
また、適応則入力DIDadpに関しては、ＳＴＥＰ２４−６で制御サイクル毎に求められる切換関数σ2 の値を制御サイクル毎に累積加算することで、切換関数σ2 の積算値Σσ2 を求める。そして、この積算値Σσ2 を用いて前記式（４４）の演算を行うことで、適応則入力DIDadpを求める。この場合において式（４４）中の係数パラメータｓ3 は「１」であり、また、式（４４）中の係数Ｆ4 は前記式（４６）の条件を満たすようにあらかじめ設定した所定値である。
【０４１２】
上記のようにして等価制御入力DIGeq 、到達則入力DIGrch、及び適応則入力DIDadpを求めた後、点火時期操作手段２６は、式（４１）に従ってそれらの総和を演算することで、今回の制御サイクルにおける点火時期偏差指令値DIG(k)を求め（ＳＴＥＰ２４−８）、図２３のルーチン処理に復帰する。
【０４１３】
図２３の説明に戻って、上記のようにしてＳＴＥＰ２３−１２で点火時期偏差指令値DIG(k)を算出した後、点火時期操作手段２６は、該点火時期偏差指令値DIG(k)の値を所定の上限値と下限値との間の値に制限する（DIG(k)＞上限値、あるいはDIG(k)＜下限値であるとき、それぞれDIG(k)の値を強制的に上限値、下限値に設定する）リミット処理を行った後（ＳＴＥＰ２３−１３）、図２２のルーチン処理に復帰する。
【０４１４】
前記ＳＴＥＰ２３−９でf/fpause＝１の場合、すなわち、ＦＩＲＥモードの中断動作を行うべき状態となっている場合には、点火時期操作回転数Ｆ／Ｂ制御を中断し、点火時期の指令値iglog を基本指令値igbaseに徐々に戻す（点火時期偏差指令値DIG を徐々に「０」に戻す）ために、その指令値iglog の制御サイクル毎の戻し量を規定する単位値dec/ig（＞０。以下、点火時期戻し単位値dec/igという）を決定する（ＳＴＥＰ２３−１４）。
【０４１５】
また、ＳＴＥＰ２３−１でf/fireon＝１の場合、すなわち、ＦＩＲＥモードの動作を終了するかもしくは行わない状態となっている場合にも、点火時期の指令値iglog を基本指令値igbaseに徐々に戻すために、前記点火時期戻し単位値dec/igを決定する（ＳＴＥＰ２３−１５）。
【０４１６】
この場合、本実施形態では、ＳＴＥＰ２３−１５では、点火時期戻し単位値dec/igをあらかじめ定めた所定値（一定値）に設定するが、ＦＩＲＥモードの中断動作に際してＳＴＥＰ２３−１４で設定する点火時期戻し単位値dec/igは、現在のスロットル弁５の開度に比例させた値（スロットル弁５の開度が大きい程、点火時期戻し単位値dec/igを大きくする）に決定する。
【０４１７】
これは次の理由による。すなわち、ＦＩＲＥモードの中断動作は、基本的には、車両のアクセルペダルの操作によって、車両の走行や内燃機関１のからぶかしを行う状況で行われるものである。そして、このときスロットル弁５の開度が前述の如くアクセル操作量Ａｐに応じた開度に操作され、該スロットル弁５の開度が大きい状況では、内燃機関１の所要の動作性能を確保するために、点火時期をなるべくすみやかに通常的な点火時期（これは基本指令値igbaseに相当する）を戻すことが好ましいと考えられる。このために、ＳＴＥＰ２３−１４では、点火時期戻し単位値dec/igをスロットル弁５の開度に比例させた値に決定する。
【０４１８】
尚、この場合に点火時期戻し単位値dec/igを決定するために必要となるスロットル弁５の開度は、コントローラ２からスロットル弁アクチュエータ２３に与える開度指令値あるいは図示しないセンサによる開度の検出値を用る。
【０４１９】
このようにして、点火時期戻し単位値dec/igを決定した後、点火時期操作手段２６は、次に、現在の点火時期偏差指令値DIG の値（これは前回の制御サイクルで決定された点火時期偏差指令値DIG(k-1)である）が「０」より小さいか否か、すなわち、遅角側の値であるか否かを判断する（ＳＴＥＰ２４−１６）。
【０４２０】
そして、DIG(k-1)＜０である場合（DIG(k-1)が遅角側の値である場合）には、現在の点火時期偏差指令値DIG(k-1)に前記ＳＴＥＰ２４−１４又は２４−１５で決定した点火時期戻し単位値dec/igを加算した値（DIG(k-1)＋dec/ig）を今回の制御サイクルにおける点火時期偏差指令値DIG(k)として決定し（ＳＴＥＰ２４−１７）、図２２のルーチン処理に復帰する。尚、この場合において、点火時期偏差指令値DIG(k)の上限値を「０」とし、（DIG(k-1)＋dec/ig）が「０」より大きくなるような場合には、点火時期偏差指令値DIG(k)の値を強制的に「０」とする。
【０４２１】
一方、ＳＴＥＰ２４−１６でDIG(k-1)≧０の場合（実質上は、DIG(k-1)＝０の場合）には、ＦＩＲＥ実行可否フラグf/fireonの値を判断し（ＳＴＥＰ２４−１８）、f/fireon＝１の場合（この場合は、ＦＩＲＥモードの中断動作中である）には、今回の制御サイクルにおける点火時期偏差指令値DIG(k)を「０」として（ＳＴＥＰ２４−１９）、図２２のルーチン処理に復帰する。
【０４２２】
尚、ＦＩＲＥモードの中断動作中（f/fireon＝１且つf/fpauseの状態）は、切換関数σ2 の積算値Σσ2 の値は中断動作の開始直前の値に保持される。
【０４２３】
また、ＳＴＥＰ２４−１８でf/fireon＝０の場合、すなわち、ＦＩＲＥモードの終了状態である場合には、ＦＩＲＥ中断フラグf/fpause及び回転数Ｆ／Ｂ実行可否フラグf/nefbの値を「０」にリセットすると共に、点火時期偏差指令値DIG(k)、切換関数σ2 の値及びその積算値Σσ2 、等価制御入力DIGeq 、到達則入力DIGrch、適応則入力DIGadpの値等を「０」に初期化した後（ＳＴＥＰ２４−２０）、図２２のルーチン処理に復帰する。
【０４２４】
図２２の説明に戻って、前述の如く、ＳＴＥＰ２２−２で点火時期偏差指令値DIG を決定する処理を実行した後、点火時期操作手段２６は、ＳＴＥＰ２２−１で決定した基本指令値igbaseにＳＴＥＰ２２−２で決定した点火時期偏差指令値DIG を加算することで、今回の制御サイクルにおける点火時期の指令値iglog を求める（ＳＴＥＰ２２−３）。
【０４２５】
次いで、点火時期操作手段２６は、前述した吸入空気量操作手段２５の前記点火時期応動補正処理に関して図１２示した点火時期の遅角側の限界値IGLGG とこの限界値IGLGG よりも若干進角側の閾値IGX とを決定する（ＳＴＥＰ２２−４）。
【０４２６】
この場合、点火時期の遅角側の限界値IGLGG は、内燃機関１の機関温度Ｔｗから図示しないデータテーブルに基づいて求められ、該限界値IGLGG 以上の進角側の点火時期では、内燃機関１の正常な運転を行い得るように定められる。また、閾値IGX は、限界値IGLGG にあらかじめ定めた所定値（一定値）を加算した値（限界値IGLGG よりも所定値だけ進角側の値）に設定する。尚、限界値IGLGG は、内燃機関１の機関温度Ｔｗが通常的な温度範囲にある場合には、一定値であるが、機関温度Ｔｗがかなり低温な温度範囲にある場合には、通常的な温度範囲の場合よりも進角側の値になるようにしている。
【０４２７】
このようにして点火時期の遅角側の限界値IGLGG と閾値IGX とを決定した後、点火時期操作手段２６は、前記ＳＴＥＰ２２−３で求めた指令値iglog を閾値IGX と比較する（ＳＴＥＰ２２−５）。そして、このとき、指令値iglog が閾値IGX に等しいかもしくは該閾値IGX よりも進角側の値である場合（iglog ≧IGX ）には、前記点火時期応動補正処理を行わないため前記点火時期判別フラグf/dec （図１５のＳＴＥＰ１５−７を参照）の値を「０」に設定し（ＳＴＥＰ２２−６）、ＳＴＥＰ２２−９に進む。
【０４２８】
また、ＳＴＥＰ２２−５でiglog ＜IGX である場合には、前記点火時期応動補正処理を行うべく前記点火時期判別フラグf/dec の値を「１」に設定し（ＳＴＥＰ２２−７）、さらに、吸入空気量操作手段２５による前記学習補正処理に係わる前記学習演算終了フラグf/flrnend の値を判断する（ＳＴＥＰ２２−８）。そして、点火時期操作手段２６は、f/flrnend ＝０である場合にのみ、現在のＦＩＲＥ経過時間t/fireの値を前記学習終了時刻パラメータt/kil の値として保持しておき（ＳＴＥＰ２２−９）、さらに前記基本学習補正係数vpskisldの算出処理を終了すべく学習演算終了フラグf/flrnend の値を「１」に設定する（ＳＴＥＰ２２−１０）。
【０４２９】
次いで、点火時期操作手段２６は、ＳＴＥＰ２２−３で求めた指令値iglog をＳＴＥＰ２２−４で決定した遅角側の限界値IGLGG と比較し（ＳＴＥＰ２２−１１）、iglog ≧IGLGG である場合、すなわち指令値iglog が遅角側の限界値IGLGG 内に収まっている場合には、図４のメインルーチンの処理に復帰する。
【０４３０】
一方、iglog ＜IGLGG である場合、すなわち、ＳＴＥＰ２２−３で求めた指令値iglog が、遅角側の限界値IGLGG を超えて遅角側の値となっている場合には、指令値iglog を強制的に限界値IGLGG に設定する（ＳＴＥＰ２２−１２）。さらにこのとき、切換関数σ2 の値の積算値Σσ2 の値をiglog ＜IGLGG となる直前の制御サイクルで求めた値に強制的にホールド（保持）した後（ＳＴＥＰ２２−１３）、図４のメインルーチンの処理に復帰する。このように積算値Σσ2 の値をホールドするのは、点火時期の指令値iglog を強制的に限界値IGLGG に設定する状態で、前記ＳＴＥＰ２３−１２の処理により点火時期偏差指令値DIG の算出を継続して行うと、切換関数σ2 の値の積算値Σσ2 の値、ひいては前記適応則入力DIGadpの値（絶対値）が過剰に大きくなってしまうからである。
【０４３１】
以上、図１９〜図２４を参照して説明した処理が、前記図４のＳＴＥＰ４−６で点火時期の指令値iglog を制御サイクル毎に生成する処理の詳細であり、このようにして決定された指令値iglog は前記点火装置２１に与えられる。そして、該点火装置２１は、与えられた指令値iglog に従って内燃機関１の点火時期を操作する。
【０４３２】
以上説明した本実施形態のシステムの作動によって、前記ＦＩＲＥモードの動作では、バイパス開度の操作により内燃機関１の燃焼室４の吸入空気量が通常のアイドリング運転時よりも増量される。また、これと並行して、点火時期操作回転数Ｆ／Ｂ制御によって、内燃機関１の回転数Ｎｅを所要の目標回転数ne/fire （最終的にはアイドリング回転数NOBJ）に収束させるように点火時期が遅角側に操作される。これにより、内燃機関１が混合気の燃焼により生成する排ガスによって触媒装置３に与えられる熱量が通常のアイドリング運転時よりも多くなって、該触媒装置３の昇温・活性化を早期に行うことが可能となると同時に、吸入空気量の増量によって上昇傾向となる内燃機関１の回転数Ｎｅを適正な回転数に維持することができる。
【０４３３】
この場合、吸入空気量の増量に関しては、その増量を開始してから触媒装置３に実際に与えられる瞬時瞬時の熱量の積算値に相当する前記推定積算吸入空気量qair/preを、触媒装置３に実際に与えるべき熱量の目標積算値に相当する目標積算吸入空気量qair/cmdに収束させるように、吸気側適応ＳＬＤ制御を用いた前記吸気量Ｆ／Ｂ制御補正処理によってバイパス開度の指令値が補正される。
【０４３４】
例えば図２５の中段に示す如く、前記構造要因による吸入空気量のばらつきに起因して、推定積算吸入空気量qair/preが標準開度指令値θ0 に応じて定まる目標積算吸入空気量qair/cmdに対して誤差（吸気偏差Ｅｑ）を生じたとき、開度指令Θ（バイパス開度の指令値）は、同図の上段に示す如く、標準開度指令θ0 に対して前記ＳＬＤ開度補正量i/sld だけ補正される。そして、この補正により、同図の中段に示す如く推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させ、ひいては、触媒装置３に実際に与えられる熱量の積算値をその目標値に追従させることができる。これにより、前記構造要因による吸入空気量のばらつきを補償し、ひいては、該構造要因による触媒装置３の昇温形態のばらつきを解消することができる。
【０４３５】
さらに、本実施形態では、前記大気条件補正処理により求めた前記大気圧補正係数kpa と大気温度補正係数kta とによって、開度指令Θを乗算補正する。すなわち、大気圧Ｐａが低い程、開度指令Θを増加側に補正し、また、大気温度が高い程、開度指令Θを増加側に補正する。これにより、大気条件による吸入空気量のばらつきを補償し、ひいては、該大気条件による触媒装置３の昇温形態のばらつきも解消することができる。
【０４３６】
この結果、ＦＩＲＥモードにおける触媒装置３の昇温形態を所望の昇温形態に合わせることができ（本実施形態では、内燃機関１の始動時の機関温度Ｔｗや、ＦＩＲＥモードの動作中の自動変速機のシフト位置が一定であれば、ＦＩＲＥモードの中断動作が行われる場合を除いて、各回のＦＩＲＥモードの動作による触媒装置３の昇温形態はほぼ同じになる）、該ＦＩＲＥモードの動作による触媒装置３の所望の昇温・活性化を確実に行うことができる。
【０４３７】
また、前記標準開度指令値θ0 、ひいては目標吸入空気量gair/cmd及び目標積算吸入空気量qair/cmdは、内燃機関１の始動時における触媒装置３の温度状態に相当する内燃機関１の始動時の機関温度Ｔｗに応じたものに設定するため（基本的には、該機関温度Ｔｗが高い程、目標吸入空気量gair/cmd及び目標積算吸入空気量qair/cmdは小さくなる）、内燃機関１の始動時の触媒装置３の温度状態に適した該触媒装置３の昇温・活性化を行うことができる。つまり、ＦＩＲＥモードにおける触媒装置３の昇温形態（経時的な温度の上昇度合い）は、内燃機関１の始動時における触媒装置３の温度状態に応じたものとなるが、ＦＩＲＥモードの動作による最終的な触媒装置３の温度状態を該触媒装置３の活性化の上で適正な温度状態にすることができる。
【０４３８】
また、標準開度指令値θ0 により規定される目標積算吸入空気量qair/cmdは、それに対応して内燃機関１の燃焼室４に制御サイクル毎に吸入されるべき吸入空気量、すなわち目標吸入空気量gair/cmdを、ＦＩＲＥ経過時間t/fireが所定値ｔ2 （図７を参照）に達した後（内燃機関１の暖機がある程度進行した後）は、緩やかに減少させていくように設定される。このため、内燃機関１の暖機の進行によって、その各部のフリクションが低下しても内燃機関１の回転数Ｎｅが上昇傾向となるのを予防することができる。そして、この結果、内燃機関１の点火時期が前記点火時期操作Ｆ／Ｂ制御によって、過剰に遅角側に操作されるのを予防することができる。
【０４３９】
さらに、本実施形態では、点火時期操作回転数Ｆ／Ｂ制御によって決定される点火時期の指令値iglog が、該点火時期を操作し得る遅角側の限界値IGLGG に近い閾値IGX を超えたときには、前記吸気量Ｆ／Ｂ制御補正処理を中断して、前記点火時期応動補正処理によって、開度指令Θをフィードフォワード的に減少させる。このため、内燃機関１のフリクションの低下が予想以上に大きく、あるいは早期に生じた場合であっても、内燃機関１の回転数Ｎｅの上昇傾向の高まりを抑え、点火時期が遅角側の限界値IGLGG に達する程、過剰に遅角側に操作されるような事態を回避することができる。
【０４４０】
また、本実施形態では、前記吸気量Ｆ／Ｂ制御補正処理に、外乱や制御対象のモデル化誤差や外乱の影響を受けにくいスライディングモード制御を用い、特に、外乱等の影響を極力排除するための適応則を用いた適応スライディングモード制御（吸気側適応ＳＬＤ制御）を用いることで、推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させ、ひいては、触媒装置３に与える熱量の積算値をその目標値に収束させる制御を高い安定性で行うことができる。この結果、触媒装置３の所望の昇温・活性化をより確実に行うことができる。
【０４４１】
また、前記吸気量Ｆ／Ｂ制御補正処理の吸気側適応ＳＬＤ制御にあっては、その制御対象を開度指令Θから積算吸入空気量Ｑａを生成する系として、離散系のモデル（吸気制御対象モデル）で表現したため、該制御対象を連続系のモデルで表現する場合に比して、点火時期側適応ＳＬＤ制御のアルゴリズムを簡素でコンピュータ処理に適したものに構築することができる。
【０４４２】
特に、吸気側適応ＳＬＤ制御の制御対象を離散系でモデル化することで、吸気側適応ＳＬＤ制御に用いる切換関数σ1 を、前記吸気偏差Ｅｑの変化速度等を用いることなく、該吸気偏差Ｅｑのみの時系列データを用いて構成することができる。この結果、前記ＳＬＤ開度補正量i/sld を求めるために必要な切換関数σ1 の値の信頼性を高め、ひいては、吸気側適応ＳＬＤ制御の信頼性を高めることができる。
【０４４３】
さらに、吸気側適応ＳＬＤ制御では、等価制御入力Θeqのフィードバック項Θeq/fb を省略して、フィードフォワード項Θeq/ff に相当する標準開度指令値θ0 を本実施形態での等価制御入力Θeqとすることで、制御対象のモデル（吸気制御対象モデル）のモデルパラメータａ1 ，ａ2 ，ｂ1 を用いずに簡略なアルゴリズムで、ＳＬＤ開度補正量i/sld を求めることができる。
【０４４４】
尚、ＳＬＤ開度補正量i/sld をモデルパラメータａ1 ，ａ2 ，ｂ1 を用いて求めるようにしてもよいことはもちろんであるが、この場合であっても、吸気側適応ＳＬＤ制御の制御対象のモデルを離散系で構築しているため、既知の同定アルゴリズムを用いることで、モデルパラメータａ1 ，ａ2 ，ｂ1 の値を的確に同定することが可能である。
【０４４５】
また、前記吸気量Ｆ／Ｂ制御補正処理にあっては、図２５の中段に示される如く、内燃機関１の始動直後の前記ＳＬＤ補正制限時間TISLDLMTが経過するまでは、推定積算吸入空気量qair/pre及び目標積算吸入空気量qair/cmdの値を強制的に「０」に保持し、ひいては、ＳＬＤ開度補正量i/sld の値も「０」に保持する。
【０４４６】
このため、内燃機関１の始動直後は、ＳＬＤ補正制限時間TISLDLMTが経過するまで吸気量Ｆ／Ｂ制御補正処理が行われないこととなり、バイパス開度は、標準開度指令値θ0 （より正確には、標準開度指令値θ0 に大気圧補正係数kpa 、大気温度補正係数kta 及び学習補正係数kilearn を乗算した値）を主体として、フィードフォワード的に操作されることとなる（図２５の上段を参照）。
【０４４７】
従って、内燃機関１の始動直後は、吸入空気量は、標準開度指令値θ0 の上昇形態と同じような形態で滑らかに増量されていくこととなり、この結果、始動直後の内燃機関１の燃焼状態を円滑に安定化し、内燃機関１の良好なエミッション状態を確保することができる。
【０４４８】
さらに、吸気量Ｆ／Ｂ制御補正処理を実際に開始する際には、図２５の下段に示す如く前記吸気側適応ＳＬＤ制御に係わるポールpole/iの値を、ＦＩＲＥ経過時間t/fireが所定値TPOLEXに達するまで、「−１」側から徐々に増加させる（｜pole/i｜を徐々に減少させる）ことで、前記吸気偏差Ｅｑの減衰速度をＦＩＲＥ経過時間t/fireが所定値TPOLEXに達するまでは、該所定値TPOLEXに達した後よりも遅くする。
【０４４９】
この結果、吸気量Ｆ／Ｂ制御補正処理の開始によって、開度指令Θ、ひいては、吸入空気量が急激に大きく変化するような事態が回避され、これによっても、始動後の初期段階における内燃機関１の燃焼状態の安定性を保ちつつ、内燃機関１の良好なエミッション状態を確保することができる。
【０４５０】
また、本実施形態では、前記学習補正処理によって、ＦＩＲＥモードの動作毎に吸気量Ｆ／Ｂ制御補正処理によるＳＬＤ開度補正量i/sld を学習し、該ＳＬＤ開度補正量i/sld をＦＩＲＥモードの動作中に積算してなるＳＬＤ積算吸気補正量qair/sldの、目標積算吸入空気量qair/cmdに対する比の値に基づく学習補正係数kilearn （基本的には前記基本学習補正係数vpskisld）を求める。そして、次回のＦＩＲＥモードの動作に際して、この学習補正係数kilearn によりＦＩＲＥモードの全期間にわたって開度指令Θをフィードフォワード的に乗算補正することで、ＦＩＲＥモード中のＳＬＤ開度補正量i/sld による制御サイクル毎の開度指令Θの補正を最小限に留めることができる。この結果、構造要因による吸入空気量のばらつきが大きく生じたような場合であっても、吸入空気量の時間的変化のパターンが標準開度指令値θ0 のパターンに対して大きく逸脱した変化を呈するような事態を回避することができ、内燃機関１の燃焼状態やエミッション状態を損なうことなく安定した内燃機関１の運転を行うことができる。
【０４５１】
また、本実施形態では、ＦＩＲＥモードの動作中に車両のアクセルペダルの操作による内燃機関１のからぶかしや車両の発進・走行が行われて、内燃機関１のアイドリング運転以外の運転を行う状態となり、ＦＩＲＥモードの中断動作を行う状況でも、吸入空気量の増量側へのバイパス開度の操作を行う（但し、吸気量Ｆ／Ｂ制御補正処理は中断）と共に、推定積算吸入空気量qair/pre及び目標積算吸入空気量qair/cmdの算出を継続する。そして、該中断動作の解除後に、吸気量Ｆ／Ｂ制御補正処理を再開して、推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させるように開度指令Θを補正する。このため、ＦＩＲＥモードの動作中に、内燃機関１のアイドリング運転以外の運転を行う状況となり、ＦＩＲＥモードの中断動作を行うような状況が生じても、ＦＩＲＥモード制限時間TFIRELMT内における触媒装置３の昇温・活性化を確実に行うことができる。
【０４５２】
また、点火時期操作回転数Ｆ／Ｂ制御に関しては、そのフィードバック制御処理に、前記吸気量Ｆ／Ｂ制御補正処理の場合と同様に、適応スライディングモード制御（点火時期側適応ＳＬＤ制御）を用いることで、内燃機関１の回転数Ｎｅの目標回転数ne/fire への制御の安定性を高めることができる。
【０４５３】
この場合、特に、応答指定型制御である点火時期側適応ＳＬＤ制御の応答指定特性を利用し、回転数Ｎｅの制御のために操作する点火時期の指令値iglog が遅角側の値である程、回転数偏差Ｅｎ（＝Ｎｅ−ne/fire ）の減衰速度を遅くするように、該減衰速度を規定するパラメータである前記ポールpole/ig （＝ｓ4 ／ｓ3 ）の通常的な値として設定される前記点火時期対応ポール基本値pole/igtblを可変的に定める。つまり、一般に、点火時期が遅角側である程、該点火時期の変化に対する内燃機関１の回転数Ｎｅの変化が大きなものとなるが、このような状態では、回転数偏差Ｅｎの減衰速度を遅めにするようにポールpole/ig の値を設定することで、点火時期側適応ＳＬＤ制御により決定される指令値iglog の変化を抑制し、回転数Ｎｅが目標回転数ne/fire に対して急変するのを回避する。また、逆に点火時期の進角側では、回転数偏差Ｅｎの減衰速度を速めにするようにポールpole/ig の値を設定することで、回転数Ｎｅの目標回転数ne/fire への迅速な追従性を確保する。これにより、回転数Ｎｅの目標回転数ne/fire への制御の適正な速応性を確保しつつ安定性を高めることができる。
【０４５４】
また、点火時期操作回転数Ｆ／Ｂ制御の開始直後の初期段階（回転数Ｆ／Ｂ経過時間Δt/nfb が所定値T/NFBXに達するまで）では、回転数偏差Ｅｎの減衰速度を遅くするように前記時間対応補正係数kigtにより点火時期対応ポール基本値pole/igtblを乗算補正してポールpole/ig を決定する。これにより、点火時期操作回転数Ｆ／Ｂ制御の開始直後に回転数Ｎｅが急激に変動し、内燃機関１の燃焼状態が悪化するような事態を回避することができる。
【０４５５】
さらに、本実施形態では、ＦＩＲＥモードの中断動作の解除後に点火時期操作回転数Ｆ／Ｂ制御を再開するに際して、内燃機関１の回転数Ｎｅが目標回転数ne/fire に対して高回転側に大きく離間している程、回転数偏差Ｅｎの減衰速度を遅くするように前記回転数対応修正補正係数kignefにより点火時期対応ポール基本値pole/igtblを乗算補正してポールpole/ig を決定する。これにより、点火時期操作回転数Ｆ／Ｂ制御を再開に際して内燃機関１の回転数Ｎｅが目標回転数ne/fire に対して高回転側に大きく離間しているときに、大きな回転数偏差Ｅｎに応じて急激に点火時期の指令値iglog が遅角側に変化するのを抑制し、内燃機関１の安定した運転状態を確保することができる。
【０４５６】
さらに本実施形態では、点火時期操作回転数Ｆ／Ｂ制御に係わる前記回転数偏差Ｅｎの減衰速度を規定するポールpole/ig の絶対値よりも、前記吸気量Ｆ／Ｂ制御補正処理に係わる前記吸気偏差Ｅｑの減衰速度を規定するポールpole/iの絶対値の方が大きな値に設定される（図１７のＳＴＥＰ１７−７、１７−８の処理を参照）。換言すれば、ポールpole/ig は、回転数偏差Ｅｎの減衰速度が吸気偏差Ｅｑの減衰速度よりも速くなるような値に設定される。このため、点火時期操作回転数Ｆ／Ｂ制御により内燃機関１の回転数Ｎｅを目標回転数ne/fire に収束させるフィードバック制御と、吸気量Ｆ／Ｂ制御補正処理により推定積算吸入空気量qair/preを目標積算吸入空気量qair/cmdに収束させるフィードバック制御との相互の干渉を回避して内燃機関１の回転数Ｎｅを安定して目標回転数ne/fire に制御することができる。
【０４５７】
また、本実施形態では、点火時期操作回転数Ｆ／Ｂ制御に用いる点火時期側適応ＳＬＤ制御では、その制御対象を点火時期の指令値iglog に相当する点火時期偏差指令値DIG から、内燃機関１の回転数Ｎｅに相当する偏差回転数DNE を生成する系として、離散系のモデル（回転数制御対象モデル）で表現したことで、点火時期側適応ＳＬＤ制御のアルゴリズムを簡素でコンピュータ処理に適したものとすることができる。
【０４５８】
特に、点火時期側適応ＳＬＤ制御の制御対象を離散系でモデル化することで、吸気側適応ＳＬＤ制御の場合と同様、切換関数σ2 を、前記回転数偏差Ｅｎの変化速度等を用いることなく、該回転数偏差Ｅｎのみの時系列データを用いて構成することができる。この結果、前記点火時期偏差指令値DIG を求めるために必要な切換関数σ2 の値の信頼性を高め、ひいては、点火時期側適応ＳＬＤ制御の信頼性を高めることができる。
【０４５９】
尚、本実施形態では、ＦＩＲＥモードにおける内燃機関１の燃焼室４の吸入空気量の増量のための流量制御弁としてバイパス弁７を用いたが、スロットル弁５の開度を操作することで、吸入空気量の増量を行うようにしてもよい。
【０４６０】
また、前述の実施形態では、燃焼室４で燃焼させる混合気の空燃比が一定であるとした場合に内燃機関１の発生熱量（排ガスの熱量）、ひいては触媒装置３に与えられる熱量が燃焼室４の吸入空気量にほぼ比例することから触媒装置３に与える熱量を表す熱量データとして制御サイクル毎の吸入空気量の積算値を用いている。しかし、内燃機関１の発生熱量は、内燃機関１の点火時期によっても若干変化することから、触媒装置３に与える熱量を表す熱量データの精度をより高める必要がある場合には、瞬時瞬時（制御サイクル毎）の吸入空気量の推定値あるいは検出値を、その時々の点火時期に応じて補正し、それを積算することで、熱量データを取得するようにしてもよい。また、混合気の空燃比を変化させる必要が生じた場合には、点火時期に応じた補正と同様に、瞬時瞬時（制御サイクル毎）の吸入空気量の推定値あるいは検出値を、その時々の空燃比に応じて補正し、それを積算することで、熱量データを取得するようにしてもよい。いずれの場合にあっても、点火時期や空燃比に応じた補正を行うための補正係数をあらかじめ設定したデータテーブル等を用意しておくことで対応することが可能である。
【０４６１】
尚、触媒装置に与える熱量を表す熱量データを取得するに際しての上記のような補正は、該熱量データとして燃料供給量等、その他のパラメータを用いる場合においても同様に適用可能である。
【図面の簡単な説明】
【図１】本発明の一実施形態を含む内燃機関の制御システムの全体構成図。
【図２】図１のシステムの内燃機関の吸気系を模式化して示した図。
【図３】図１のシステムの基本的作動を説明するための線図。
【図４】図１のシステムの作動を説明するためのフローチャート。
【図５】図１のシステムの作動を説明するためのフローチャート。
【図６】図１のシステムの作動を説明するための線図。
【図７】図１のシステムの作動を説明するための線図。
【図８】図１のシステムの作動を説明するための線図。
【図９】図１のシステムの作動を説明するための線図。
【図１０】図１のシステムの作動を説明するための線図。
【図１１】図１のシステムの作動を説明するための線図。
【図１２】図１のシステムの作動を説明するための線図。
【図１３】図１のシステムの作動を説明するための線図。
【図１４】図１のシステムの作動を説明するためのフローチャート。
【図１５】図１のシステムの作動を説明するためのフローチャート。
【図１６】図１のシステムの作動を説明するためのフローチャート。
【図１７】図１のシステムの作動を説明するためのフローチャート。
【図１８】図１のシステムの作動を説明するためのフローチャート。
【図１９】図１のシステムの作動を説明するための線図。
【図２０】図１のシステムの作動を説明するための線図。
【図２１】図１のシステムの作動を説明するための線図。
【図２２】図１のシステムの作動を説明するためのフローチャート。
【図２３】図１のシステムの作動を説明するためのフローチャート。
【図２４】図１のシステムの作動を説明するためのフローチャート。
【図２５】図１のシステムの作動を説明するための線図。
【符号の説明】
１…内燃機関、３…触媒装置、７…バイパス弁（流量制御弁）、８…バイパス通路（吸入空気通路）、２５…吸入空気量操作手段、２６…点火時期操作手段。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a control device for an internal combustion engine.
[0002]
[Prior art]
In general, a catalyst device such as a three-way catalyst provided in an exhaust system of an internal combustion engine for purifying exhaust gas of the internal combustion engine requires the exhaust gas if its temperature rises to some extent and the catalyst device is not activated. It is difficult to ensure the purification performance. For this reason, ensuring the purification performance of the exhaust gas immediately after the start of the internal combustion engine in which the catalyst device is relatively low temperature has been an important issue.
[0003]
The applicant of the present application has proposed a technique for solving such a problem in Japanese Patent Application No. 9-109197 or Japanese Patent Application No. 10-10326, and the outline of this technique is as follows.
[0004]
That is, at the time of the first idling operation after starting the internal combustion engine, the amount of intake air sucked into the combustion chamber of the internal combustion engine is set to be larger than that during normal idling operation (for example, idling operation after driving an automobile equipped with the internal combustion engine). In order to increase the amount, the opening degree of a flow control valve (specifically, a throttle valve or a control valve of a bypass passage bypassing the throttle valve) provided in the intake air passage of the internal combustion engine is manipulated. Further, after the start of the increase of the intake air amount, the rotational speed (actual rotational speed) of the internal combustion engine that tends to increase due to this increase is converged to the required target rotational speed for the idling operation immediately after the start of the internal combustion engine. The ignition timing of the internal combustion engine is operated by feedback control, whereby the ignition timing is corrected to the retard side than in the normal case.
[0005]
By increasing the intake air amount of the internal combustion engine and operating the ignition timing toward the retarded angle side, the amount of heat of the exhaust gas generated by the internal combustion engine increases, and as a result, the catalyst device that is warmed by the exhaust gas becomes early. Activated. As a result, the required purification performance of the catalyst device can be ensured at an early stage after the internal combustion engine is started.
[0006]
Also, with this technology, the rotational speed of the internal combustion engine that tends to increase as a result of the increase in the intake air amount of the internal combustion engine is feedback-controlled to the target rotational speed, so that the rotational speed of the internal combustion engine is stably maintained as a result. However, since the ignition timing is operated to the retard side, the intake air amount operation and the ignition timing operation can be performed independently of each other. For this reason, the construction of these control systems can be facilitated and the control system can be simplified.
[0007]
By the way, in such a technique, the opening degree of the flow rate control valve during the increase of the intake air amount is determined so that the operation after the start of the internal combustion engine is performed in a smooth and good combustion state as much as possible. It is set so that the amount of heat that can sufficiently activate the catalyst device in time can be given to the catalyst device. In this case, the actual intake air amount of the internal combustion engine is uniquely determined by the opening of the flow control valve, and the size of the command value of the opening and the form of temporal change are determined in advance. The opening degree of the flow control valve was operated according to the command value.
[0008]
However, the actual intake air amount of the internal combustion engine is affected by the atmospheric pressure even when the opening degree of the flow control valve is constant, and the actual intake air amount of the internal combustion engine increases as the atmospheric pressure increases. . Furthermore, although the influence of atmospheric pressure is smaller, the actual intake air amount of the internal combustion engine is also affected by the atmospheric temperature, and the higher the atmospheric temperature, the lower the intake air density. The amount of air is reduced. For this reason, the amount of heat of the exhaust gas generated by the internal combustion engine according to the amount of intake air is also affected by the atmospheric pressure and the atmospheric temperature. As a result, in an environment where the atmospheric pressure and the atmospheric temperature are different, when the opening degree of the flow control valve is operated as described above regardless of the atmospheric pressure and the atmospheric temperature, the amount of heat given to the catalyst device varies, and as a result Variations in the temperature rise and activation of the catalyst device may occur, and in some cases, the temperature rise and activation of the catalyst device may be insufficient, and sufficient purification performance may not be ensured.
[0009]
[Problems to be solved by the invention]
In view of such a background, the present invention compensates for variations in the amount of intake air of the internal combustion engine due to atmospheric pressure and atmospheric temperature, and hence the amount of heat applied to the catalytic device, and achieves a desired temperature increase of the catalytic device regardless of atmospheric pressure or atmospheric temperature. An object of the present invention is to provide a control device for an internal combustion engine that can be activated.
[0010]
[Means for Solving the Problems]
In order to achieve this object, the control device for an internal combustion engine of the present invention is an internal combustion engine that releases exhaust gas through a catalyst device. During idling operation in an operation mode for early activation of the catalytic device after startup, The intake air volume of the internal combustion engine More than the amount of intake air during idling operation in modes other than the operation mode Intake air amount operation means for operating the opening of a flow control valve provided in the intake air passage of the internal combustion engine so as to increase the predetermined amount, and the rotation speed of the internal combustion engine during the increase of the intake air amount In an internal combustion engine control device comprising ignition timing operating means for operating the ignition timing of the internal combustion engine by feedback control so as to converge to the rotational speed and correcting the ignition timing to the retard side An increase in the intake air amount by the intake air amount operation means and an operation of the ignition timing by the ignition timing operation means in an operation mode for early activation of the catalyst device at the start of the operation mode. Before the specified time elapses from before The intake air amount operation means includes means for adjusting the opening degree of the flow rate control valve when the intake air amount is increased according to the atmospheric pressure so that the opening degree is increased as the atmospheric pressure is lower. It is characterized by that.
[0011]
According to the present invention, the opening degree of the flow control valve when the intake air amount of the internal combustion engine is increased is adjusted according to the atmospheric pressure so that the opening degree is increased as the atmospheric pressure is lower. It is possible to compensate for variations in the amount of intake air of the internal combustion engine due to atmospheric pressure, and in turn, the amount of heat applied to the catalyst device. In other words, when the intake air amount of the internal combustion engine is increased, when the flow control valve is operated to a certain opening, the actual intake air amount decreases as the atmospheric pressure decreases, so that the flow control valve is opened as described above. By adjusting the degree according to the atmospheric pressure, the intake air amount of the internal combustion engine can be set to a desired intake air amount (an intake air amount that is increased by a predetermined amount compared to that during normal idling operation) regardless of the atmospheric pressure. it can. As a result, according to the present invention, the desired temperature increase and activation of the catalyst device can be performed regardless of the atmospheric pressure, and the required purification performance of the catalyst device can be reliably ensured.
[0012]
Further, in the present invention, means for adjusting the opening degree of the flow rate control valve when the intake air amount is increased according to the atmospheric temperature so that the opening degree is increased as the atmospheric temperature is higher. Provided in the quantity operation means.
[0013]
That is, when the intake air amount of the internal combustion engine is increased, when the flow control valve is operated at a certain opening, the actual intake air amount decreases as the atmospheric temperature increases. The degree is adjusted according to the atmospheric temperature so that the opening degree is increased as the atmospheric temperature is higher. As a result, the intake air amount of the internal combustion engine can be set to a desired intake air amount (an intake air amount that is increased by a predetermined amount compared with the normal idling operation) regardless of the atmospheric pressure as well as the atmospheric temperature. It is also possible to compensate for the intake air amount of the internal combustion engine due to the temperature, and hence the variation in the amount of heat given to the catalyst device. As a result, according to the present invention, the desired temperature rise and activation of the catalyst device can be performed more reliably regardless of not only the atmospheric pressure but also the atmospheric temperature, and thus the required purification performance of the catalyst device can be improved. It can be ensured more reliably.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention will be described with reference to FIGS.
[0015]
FIG. 1 shows an overall system configuration of an apparatus according to the present embodiment, in which 1 is an internal combustion engine and 2 is a controller responsible for operation control of the internal combustion engine 1.
[0016]
The internal combustion engine 1 is mounted as a propulsion source in a vehicle such as an automobile or a hybrid vehicle, for example, and is a catalyst device configured by using, for example, a three-way catalyst, exhaust gas generated by burning a mixture of fuel and air. 3 to the atmosphere.
[0017]
FIG. 2 schematically shows the intake system of the internal combustion engine 1. As shown in FIG. 2, in the combustion chamber 4 of the internal combustion engine 1, a main intake passage 6 provided with a throttle valve 5, a bypass valve 7 that bypasses the throttle valve 5 and is connected to the main intake passage 6. The air can be sucked in via the bypass passage 8 provided with the above. In FIG. 2, 9 is a cylinder, 10 is a piston, 11 and 12 are intake valves and exhaust valves, respectively, and 13 is a chamber.
[0018]
Returning to the description of FIG. 1, in the system of the present embodiment, as an incidental configuration for controlling the operation of the internal combustion engine 1, the rotational speed Ne (actual rotational speed) of the internal combustion engine 1 and the engine temperature Tw (for example, cooling water temperature) ), And the intake pressure which is the internal pressure of the main intake passage 6 on the downstream side of the throttle valve 5 and the bypass valve 7 (in this embodiment, the internal pressure of the chamber 13 in FIG. 2). An intake pressure sensor 16 that detects Pb, an atmospheric temperature sensor 17 and an atmospheric pressure sensor 18 that detect atmospheric temperature Ta and atmospheric pressure Pa, an accelerator pedal operation amount Ap (hereinafter referred to as an accelerator operation amount Ap) and a vehicle (not shown), and Various sensors such as an accelerator sensor 19 and a vehicle speed sensor 20 for detecting the vehicle speed V are provided.
[0019]
Further, as an incidental configuration for driving the internal combustion engine 1, an ignition device 21 that ignites an air-fuel mixture in the combustion chamber 4, a fuel supply device 22 that supplies fuel to the combustion chamber 4, the throttle valve 5, and the bypass valve 7. Are provided with a throttle valve actuator 23, a bypass valve actuator 24, and the like.
[0020]
Although illustration is omitted, in addition to the above-described configuration, a starter motor for starting the internal combustion engine 1, a power source battery for various electronic devices, and a transmission for transmitting the power of the internal combustion engine 1 to driving wheels (this embodiment) In this system, an automatic transmission) is also provided.
[0021]
The controller 2 is configured using a microcomputer, and the ignition device 21 is based on the output (detection value) data of the various sensors 14 to 20 described above, a predetermined program, a preset data value, and the like. The required operation of the internal combustion engine 1 is performed via the fuel supply device 22, the throttle valve actuator 23, the bypass valve actuator 24, and the like.
[0022]
The controller 2 operates the opening degree of the throttle valve 5 or the bypass valve 7 via the throttle valve actuator 23 and the bypass valve actuator 24, respectively, so as to control the intake air amount of the combustion chamber 4 of the internal combustion engine 1. A quantity operation means 25, an ignition timing operation means 26 for operating the ignition timing of the internal combustion engine 1 via the ignition device 21, and the like are provided as functional configurations.
[0023]
Although details of the functions of these means 25 and 26 will be described later, the ignition timing operation means 26 has a function as a rotation speed control device according to the present invention.
In the present embodiment, the bypass valve 7 and the bypass passage 8 correspond to the flow control valve and the intake air passage according to the present invention, respectively. In the present embodiment, the control cycle (control cycle) in which the controller 2 executes the control process is a crank angle cycle (so-called TDC).
[0024]
Next, the operation of the system of this embodiment will be described together with the functions of the intake air amount operation means 25 and the ignition timing operation means 26 of the controller 2.
[0025]
First, the outline of the basic operation of the system of this embodiment will be briefly described with reference to FIG. FIG. 3 is an upper diagram illustrating exemplary forms of temporal changes in the opening degree of the bypass valve 7 (hereinafter referred to as bypass opening degree), ignition timing, and rotational speed in the operation from the start of the internal combustion engine 1 to the idling operation. These are shown in the middle and lower stages.
[0026]
Referring to FIG. 3, in the system of the present embodiment, when the system is started by operating a start switch (not shown) or the like while the operation of internal combustion engine 1 is stopped, the operation of the system is first performed. An operation mode (hereinafter referred to as a start mode) is performed in which the internal combustion engine 1 is started while ranking is performed by a starter motor (not shown). In this starting mode, the bypass opening and the ignition timing are respectively operated as shown in the figure, and the rotational speed Ne of the internal combustion engine 1 varies as shown in the figure.
[0027]
In the system of the present embodiment, the opening of the throttle valve 5 during operation of the internal combustion engine 1 is in accordance with the accelerator operation amount Ap, and the accelerator pedal (not shown) is not operated. (Ap = 0 state. This state is an idling operation state of the internal combustion engine 1 while the vehicle is stopped), and the opening degree of the throttle valve 5 is “0” (the throttle valve 5 is closed). . In this state, intake of the combustion chamber 4 is performed only through the bypass passage 8. FIG. 3 relates to a case where intake of the combustion chamber 4 is performed only through the bypass passage 8 in this way.
[0028]
When a so-called complete explosion of the internal combustion engine 1 is confirmed in the start mode, the operation of the system is an operation mode (hereinafter referred to as FIRE) for early activation of the catalyst device 3 while performing idling operation of the internal combustion engine 1. Mode).
[0029]
In the FIRE mode, for example, a command value θCMD of a bypass opening larger than that in normal idling operation (in idling operation other than in the FIRE mode) is used in the form (pattern) of temporal change as shown in the upper part of FIG. Generate sequentially. Then, by operating the bypass opening via the bypass valve actuator 24 in accordance with the command value θCMD, the intake air amount of the combustion chamber 4 of the internal combustion engine 1 is increased from that during the normal idling operation.
[0030]
In this case, the operation in the FIRE mode is basically an elapsed time t / fire after the start of the operation (this is an elapsed time since the increase of the intake air amount is started. This is performed until the elapsed time t / fire) reaches a predetermined time limit TFIRELMT (hereinafter referred to as FIRE mode time limit TFIRELMT).
[0031]
The command value θCMD of the bypass opening in the FIRE mode is basically an exhaust gas having a calorific value that can appropriately perform the desired temperature increase and activation of the catalyst device 3 within the FIRE mode limit time TFIRELMT. The amount of heat of the exhaust gas is approximately proportional to the amount of intake air in the combustion chamber 4 of the internal combustion engine 1) to the catalyst device 3, and the combustion state and emission state of the internal combustion engine 1 in the FIRE mode are kept good. However, the idling operation of the internal combustion engine 1 is set so as to be performed stably and smoothly.
[0032]
As the intake air amount increases (increases the bypass opening) in the FIRE mode, the rotational speed Ne (actual rotational speed) of the internal combustion engine 1 immediately after the start of the increase is as shown by a solid line in the lower part of FIG. It rises in form. When the rotational speed Ne reaches a set rotational speed (NOBJ + NEFSLDS) higher by a predetermined amount NEFSLDS than the idling rotational speed NOBJ (constant) determined in advance as an appropriate rotational speed to be finally maintained in the FIRE mode, The ignition timing of the internal combustion engine 1 is operated as indicated by the solid line in the middle of FIG. 3 by feedback control for converging the engine speed Ne to the required target engine speed ne / fire. Hereinafter, this feedback control of the rotational speed Ne is referred to as ignition timing operation rotational speed F / B control. The ignition timing operation rotational speed F / B control is performed not only when the rotational speed Ne of the internal combustion engine 1 reaches the set rotational speed (NOBJ + NEFSLDS), but also at a predetermined value TSLDIGST that the FIRE elapsed time t / fire is determined in advance. It is also started when it reaches (see the lower part of FIG. 3).
[0033]
In this ignition timing operation rotational speed F / B control, the target rotational speed ne / fire of the internal combustion engine 1 is set in the form shown by the broken line in the lower part of FIG. The target rotational speed ne / fire is decreased from the set rotational speed (NOBJ + NEFSLDS) to the idling rotational speed NOBJ with a predetermined degree of decrease (inclination), and after reaching the idling rotational speed NOBJ, the idling rotational speed Held in the number NOBJ. The idling rotational speed NOBJ is set to a rotational speed higher than the rotational speed during normal idling operation.
[0034]
In the ignition timing operation rotational speed F / B control, the middle stage of FIG. 3 is performed by feedback control processing so that the rotational speed Ne (actual rotational speed) of the internal combustion engine 1 is converged to the target rotational speed ne / fire thus set. The ignition timing correction amount DIG (the correction amount DIG is an ignition timing deviation command value, which will be described later) as shown by the broken line in FIG. 3 is obtained, and the ignition timing basic command value igbase (one point in the middle of FIG. 3) is calculated by this correction amount DIG. The ignition timing command value iglog is determined by correcting (indicated by a chain line). Here, the basic command value igbase for the ignition timing corresponds to the command value for the ignition timing during normal operation of the internal combustion engine 1 (during operation other than the FIRE mode), and is a value on the advance side.
[0035]
In the ignition timing operation speed F / B control, the ignition timing of the internal combustion engine 1 is operated via the ignition device 21 in accordance with the command value iglog obtained by correcting the basic command value igbase as described above. Is controlled by feedback to the target rotational speed ne / fire (finally to the idling rotational speed NOBJ).
[0036]
At this time, since the rotational speed Ne of the internal combustion engine 1 tends to increase with respect to the target rotational speed ne / fire due to the increase in the intake air amount described above, the correction amount obtained by the ignition timing operation rotational speed F / B control. DIG corrects the ignition timing from the basic command value igbase to the retard side. Therefore, the ignition timing iglog obtained by correcting the basic command value igbase by this correction amount DIG (≦ 0) is the retard side as shown by the solid line in the middle stage of FIG.
[0037]
Thus, the FIRE mode performed at the time of the first idling operation after the internal combustion engine 1 is started increases the intake air amount by operating the bypass opening as described above, and delays the ignition timing by controlling the ignition timing operating rotational speed F / B. The internal combustion engine 1 mixes in the combustion chamber 4 while controlling the rotational speed Ne of the internal combustion engine 1 to the required target rotational speed ne / fire (finally the idling rotational speed NOBJ) by operating the angle side. The amount of heat of the exhaust gas generated by the combustion of is increased compared to that during normal idling operation. Then, by supplying the exhaust gas with the increased amount of heat to the catalyst device 3 in this way, the temperature rise and activation of the catalyst device 3 are accelerated, and the required exhaust gas purification performance of the catalyst device 3 is secured early. .
[0038]
The FIRE mode in which the increase of the intake air amount and the ignition timing operation rotational speed F / B control are performed in the FIRE elapsed time t /, except when the accelerator pedal of the vehicle is operated during the operation. The fire is continuously performed until the FIRE mode limit time TFIRELMT is reached, and thereafter, the operation of the system shifts to a mode in which the internal combustion engine 1 is normally operated (hereinafter referred to as a normal mode). In this normal mode, the bypass opening is manipulated, for example, to an opening for performing a normal idling operation of the internal combustion engine 1 (<bypass opening in the FIRE mode. Refer to the right side in the upper part of FIG. 3). The ignition timing of the internal combustion engine 1 is gradually returned to the normal ignition timing determined by the basic command value igbase after the end of the FIRE mode, as shown in the right part of the middle stage of FIG. .
[0039]
In the system of the present embodiment, the vehicle is started or traveled during the operation of the FIRE mode (before the FIRE elapsed time t / fire reaches the FIRE mode limit time TFIRELMT), or the internal combustion engine 1 is so-called. When an accelerator pedal (not shown) is operated in order to perform the overshoot, and an operation other than idling of the internal combustion engine 1 is to be performed, the FIRE mode is interrupted.
[0040]
In this interruption operation, the intake air amount is continuously increased by the operation of the bypass opening as described above in order to ensure the temperature rise and activation of the catalyst device 3. In order to ensure the required power performance of the engine 1, the ignition timing is returned to the normal ignition timing determined by the basic command value igbase (ignition timing operation speed F / B control is interrupted). ). Then, when the idling operation of the internal combustion engine 1 is again performed within the FIRE mode limit time TFIRELMT, the ignition timing operation rotational speed F / B control is resumed. In other words, the interruption operation in the FIRE mode is basically an operation for interrupting the ignition timing operation rotational speed F / B control. However, even when the intake air amount is increased by the operation of the bypass opening, the partial control processing related to the increase is interrupted.
[0041]
The contents described above are the outline of the basic operation of the system of this embodiment.
[0042]
Next, the detailed operation of the system of this embodiment will be described below in consideration of such basic operations.
[0043]
When the system of the present embodiment is started in a state where the operation of the internal combustion engine 1 is stopped, the controller 2 executes a main routine process shown in the flowchart of FIG. 4 in a predetermined control cycle, that is, a crank angle cycle (TDC).
[0044]
First, the controller 2 determines whether or not the operation mode of the system is the start mode (STEP 4-1). This determination is made, for example, based on whether or not a so-called complete explosion of the internal combustion engine 1 has been confirmed. The operation mode is the start mode after the system is started until the complete explosion is confirmed. The complete explosion is confirmed based on the output of the rotation speed sensor 14 (detection value of the rotation speed Ne) and the like.
[0045]
If it is determined in STEP 4-1 that the operation mode is the start mode, the controller 2 executes a start mode process for starting the internal combustion engine 1 for each control cycle (STEP 4-2).
[0046]
In this starting mode process, the controller 2 outputs the command values for the ignition timing, fuel supply amount, and bypass opening appropriate for starting the internal combustion engine 1 to the outputs (detected values) of the various sensors 14 to 20 described above. ), A predetermined map, an arithmetic expression, or the like. A starter motor (not shown) is operated while operating the ignition timing, fuel supply amount, and bypass opening (intake air amount) via the ignition device 21, the fuel supply device 22, and the bypass valve actuator 24 in accordance with the determined command value. The internal combustion engine 1 is started by cranking the internal combustion engine 1 according to the above.
[0047]
In the start mode process, the controller 2 performs initial setting of various parameters (details will be described later) such as a flag used in the control process in the FIRE mode.
[0048]
Further, in the start mode process, the engine temperature Tw, the atmospheric temperature Ta, and the atmospheric pressure Pa when the internal combustion engine 1 is started are detected by the engine temperature sensor 15, the atmospheric temperature sensor 17, and the atmospheric pressure sensor 18, respectively, and stored in a memory (not shown). Retained.
[0049]
On the other hand, if it is determined in STEP 4-1 that the operation mode is not the start mode (when the complete explosion of the internal combustion engine 1 is confirmed), the controller 2 supplies fuel to the internal combustion engine 1 every control cycle. A quantity command value is generated (STEP 4-3). Further, the controller 2 determines whether or not the control processing in the FIRE mode should be performed (whether the operation mode is set to the FIRE mode or the normal mode) (STEP 4-4), and then commands the bypass opening degree. The value θCMD is generated by the intake air amount operating means 25 (STEP 4-5), and the command value iglog of the ignition timing of the internal combustion engine 1 is generated by the ignition timing operating means 26 (STEP 4-6).
[0050]
In the fuel supply amount command value generation process in STEP 4-3, first, from the rotational speed Ne (actual rotational speed) and the intake pressure Pb of the internal combustion engine 1 detected by the rotational speed sensor 14 and the intake pressure sensor 16, A basic fuel supply amount is obtained based on a predetermined map. The basic fuel supply amount is corrected according to the engine temperature Tw, the atmospheric temperature Ta, etc. detected by the engine temperature sensor 15 and the atmospheric temperature sensor 17, so that the intake air amount of the combustion chamber 4 of the internal combustion engine 1 is obtained. A corresponding fuel supply amount command value is generated.
[0051]
The command value of the fuel supply amount thus generated is given from the controller 2 to the fuel supply device 22 every control cycle, and the fuel supply device 22 supplies fuel to the internal combustion engine 1 according to the given command value. Supply.
[0052]
Further, the condition determination process in STEP 4-4 is performed as shown in the flowchart of FIG.
[0053]
That is, the controller 2 first determines whether the current FIRE elapsed time t / fire is within the FIRE mode limit time TFIRELMT (whether t / fire <TFIRELMT, STEP 5-1), and the rotation speed. Whether or not the current rotational speed Ne detected by the sensor 14 is within a normal predetermined range (STEP5-2), the current engine temperature Tw detected by the engine temperature sensor 15 is within a normal predetermined range. Whether or not (STEP5-3). The FIRE elapsed time t / fire determined in STEP 5-1 is initialized to “0” in the start mode process (STEP 4-2), and when the start mode ends (the complete explosion of the internal combustion engine 1 is confirmed). Control cycle).
[0054]
When the conditions of these STEPs 5-1 to 5-3 are not satisfied, that is, when the FIRE elapsed time t / fire has already reached the FIRE mode limit time TFIRELMT, or the current rotational speed Ne of the internal combustion engine 1 is abnormal. When the engine speed is high or low, or when the current engine temperature Tw of the internal combustion engine 1 is abnormally high or low, the controller 2 performs processing related to learning calculation (described later) ( Specifically, the value of the flag f / flrnend (hereinafter referred to as the learning calculation end flag f / flrnend) represented by the values “1” and “0”, respectively, indicates whether or not to end the basic learning correction coefficient vpskisld calculation process described later. Judgment is made (STEP 5-4). Then, only when f / flrnend = 0, the controller 2 holds the current value of the FIRE elapsed time t / fire as the value of the parameter t / kil (STEP 5-5), and further, the learning calculation end flag The value of f / flrnend is set to “1” (STEP 5-6). Here, the parameter t / kil represents the FIRE elapsed time t / fire when the calculation processing of the basic learning correction coefficient vpskisld described later is finished. Hereinafter, the parameter t / kil is referred to as a learning end time parameter t / kil.
[0055]
Note that the values of the learning calculation end flag f / flrnend and the learning end time parameter t / kil are initialized to “0” in the start mode process (STEP 4-2).
[0056]
Next, the controller 2 forcibly fixes the value of the FIRE elapsed time t / fire to the FIRE mode limit time TFIRELMT (STEP 5-7), and then determines whether or not the above-described interruption operation of the FIRE mode should be performed. The value of the flag f / fpause (hereinafter referred to as the FIRE interruption flag f / fpause), which is represented by the values “1” and “0” respectively, and the value “1” indicating whether or not the operation of the FIRE mode should be performed ”And“ 0 ”and the value of the flag f / fireon (hereinafter referred to as FIRE execution enable / disable flag f / fireon) is set to“ 0 ”(STEP 5-8, 5-9), and the processing of the main routine of FIG. Return to. Note that the fact that the operation in the FIRE mode is not to be performed (f / fireon = 0) means that the operation in the normal mode is to be performed.
[0057]
On the other hand, when the conditions of STEPs 5-1 to 5-3 are satisfied, the controller 2 determines whether or not the accelerator pedal of the vehicle is operated based on the output of the accelerator sensor 19 (accelerator operation amount Ap) (STEP5- 10) It is sequentially determined whether or not the fuel cut of the internal combustion engine 1 is in progress (STEP 5-11). The fuel cut of the internal combustion engine 1 is a process for cutting off the fuel supply when the vehicle is decelerated. Further, in a state where the accelerator pedal of the vehicle is being operated, the controller 2 operates the throttle valve 5 through the throttle valve actuator 23 to an opening corresponding to the accelerator operation amount Ap.
[0058]
At this time, when neither of the conditions of STEP 5-10 and 5-11 is satisfied, the engine is basically in an idling operation state. In this case, the controller 2 sets the value of the FIRE interruption flag f / fpause to “0” (STEP 5-12), and further sets the value of the FIRE execution flag f / fireon to “1”. (STEP 5-13), the process returns to the main routine of FIG.
[0059]
If any of the conditions of STEP 5-10 and STEP 5-11 is satisfied, the controller 2 sets the value of the FIRE interruption flag f / fpause to “1” in order to perform the interruption operation of the FIRE mode. (STEP 5-14), the value of the learning calculation end flag f / flrnend is determined (STEP 5-15). Then, only when f / flrnend = 0, the controller 2 holds the current value of the FIRE elapsed time t / fire as the value of the learning end time parameter t / kil (STEP 5-16), and further performs a learning calculation. The value of the end flag f / flrnend is set to “1” (STEP 5-17).
[0060]
Next, the controller 2 sets the value of the countdown timer cnt / igvpl used for the ignition timing operation processing (details will be described later) after the end of the interruption operation in the FIRE mode to a predetermined initial value XCNT (STEP 5-18) Then, the processing returns to the main routine of FIG.
[0061]
The situation in which any of the conditions of STEP 5-10 and 5-11 is satisfied, that is, the situation where the value of the FIRE interruption flag f / fpause is set to “1” to perform the interruption operation of the FIRE mode is basically Is a situation where the vehicle is started and running by operating the accelerator pedal within the FIRE mode time limit TFIRELMT (a situation where the internal combustion engine 1 is driving its load), or an internal combustion engine This is a situation where 1 is being carried out. However, when the vehicle is started within the FIRE mode time limit TFIRELMT, any of the conditions of STEP 5-10 and 5-11 may not be satisfied during the subsequent deceleration of the vehicle. In other words, the situation in which the FIRE mode is interrupted is more precisely a situation in which an operation other than the idling operation of the internal combustion engine 1 is performed while the air-fuel mixture is being combusted.
[0062]
As long as the rotational speed Ne of the internal combustion engine 1 and the engine temperature Tw are within appropriate ranges after the start of the internal combustion engine 1 (after the end of the start mode), the above-described FIRE progress is achieved by the condition determination processing in STEP 4-4 described above. Until the time t / fire reaches the FIRE mode limit time TFIRELMT, the operation mode of the system is set to FIRE mode (f / fireon = 1). In the state where the FIRE mode is set, the intake air amount is increased by the operation of the bypass opening as described above, and the ignition timing operation rotational speed F except for the interruption operation of the FIRE mode. / B control is performed in parallel.
[0063]
In a situation where the rotational speed Ne of the internal combustion engine 1 and the engine temperature Tw are abnormally high or low for some reason, the system operation mode is set to the normal mode immediately after the internal combustion engine 1 is started. Alternatively, the FIRE mode is canceled (terminated) in the middle and the normal mode is set (f / fireon = 0). In the state where the normal mode is set, the bypass opening and ignition timing are operated to the bypass opening and ignition timing for performing normal operation of the internal combustion engine 1 (operation other than the FIRE mode). . Further, when the normal mode is set, in step 5-7, the value of the FIRE elapsed time t / fire is forcibly fixed to the FIRE mode limit time TFIRELMT. Thereafter, the step 5 is continued until the internal combustion engine 1 is started again. The -1 condition does not hold (the FIRE elapsed time t / fire is initialized only in the start mode). Therefore, the FIRE mode is set only in a period after the start of the internal combustion engine 1 until the FIRE mode limit time TFIRELMT elapses.
[0064]
In addition, when the system operation mode is set to the FIRE mode (f / fireon = 1), the vehicle is run by operating the accelerator pedal, or the internal combustion engine 1 is blown up. In other cases, that is, when the internal combustion engine 1 other than the idling operation should be operated (when either of the conditions of STEP 5-10 and 5-11 is satisfied), the FIRE interruption flag f / fpause is set to “1”. Is set. In this case, as described above, the FIRE mode is interrupted such that the intake air amount is increased by the operation of the bypass opening, but the ignition timing operation rotational speed F / B control is interrupted. Become. In the state where the operation mode is the FIRE mode and the interruption operation is performed (f / fireon = 1 and f / fpause = 1), any of the conditions of STEP5-10 and 5-11 is not satisfied, When the FIRE interruption flag f / fpause is returned to “0” (this is basically a case where the idling operation is resumed), the interruption operation is canceled and the ignition timing operation speed F / B control is performed. Is resumed.
[0065]
In the situation where the FIRE mode is interrupted by the operation of the accelerator pedal (STEP 5-10 condition is satisfied), the opening of the throttle valve 5 is set according to the accelerator operation amount Ap (> 0). . For this reason, in this situation, intake air is supplied to the combustion chamber 4 of the internal combustion engine 1 through both the bypass valve 7 and the throttle valve 5.
[0066]
Further, although not adopted in the present embodiment, the system operation mode may not be set to the FIRE mode until some time has elapsed after the start mode ends.
[0067]
Next, the process for generating the bypass opening command value θCMD in STEP 4-5 in FIG. 4 will be described.
[0068]
Here, before describing the specific contents of this process in detail, the basic contents of this process will be described.
[0069]
In the system of the present embodiment, as a main process performed by the intake air amount operation means 25 of the controller 2 in order to generate the bypass opening command value θCMD in the FIRE mode, the bypass opening standard command value θ0 (hereinafter referred to as standard). (Hereinafter referred to as a standard opening command value generation process) and the integrated value of the actual intake air amount of the combustion chamber 4 of the internal combustion engine 1 is converged to a required target value. Further, a process for correcting the command value of the bypass opening by the feedback control process (hereinafter referred to as an intake air amount F / B control correction process) and a correction of the command value of the bypass opening by the intake air amount F / B control correction process A process for learning the amount every time the operation in the FIRE mode is performed, and correcting the command value of the bypass opening degree in the next operation in the FIRE mode based on the learning result (hereinafter, Learning correction processing) and processing for correcting the command value of the bypass opening according to the atmospheric pressure Pa and the atmospheric temperature Ta detected by the atmospheric pressure sensor 18 and the atmospheric temperature sensor 17, respectively (hereinafter referred to as atmospheric) And a process for correcting the command value of the bypass opening according to the ignition timing operated by the ignition timing operation rotation F / B control (hereinafter referred to as an ignition timing response correction process). . The basic contents of these processes will be described below.
[0070]
First, the standard opening command value generation process will be described.
[0071]
In the system of the present embodiment, the increase of the intake air amount in the FIRE mode is mainly aimed at the early temperature rise and activation of the catalyst device 3, and in this case, the catalyst device 3 within the FIRE mode limit time TFIRELMT. So that the desired temperature increase and activation of the exhaust gas can be reliably performed (the amount of heat of the exhaust gas is approximately proportional to the amount of intake air in the combustion chamber 4 of the internal combustion engine 1). It is necessary to increase the amount of intake air.
[0072]
On the other hand, the increase in the intake air amount is started immediately after the internal combustion engine 1 is started, and the ignition timing is operated more retarded than the normal ignition timing by the ignition timing operation rotational speed F / B control. However, if the amount of intake air is increased (the pattern of change in increase over time) is inappropriate, the combustion state and emission state of the internal combustion engine 1 may be impaired. Therefore, it is necessary to increase the intake air amount in the FIRE mode so that the idling operation of the internal combustion engine 1 can be performed stably and smoothly without impairing the combustion state or emission state of the internal combustion engine 1 in the FIRE mode. There is.
[0073]
Therefore, in the standard opening command value generation process, the standard opening command value θ0, which is the basis of the bypass opening command value to be given to the bypass valve actuator 24 in order to increase the intake air amount, It is generated in a feed-forward manner for each control cycle (TDC) according to the engine temperature Tw at the time of starting the engine 1 (starting mode), the FIRE elapsed time t / fire, and the like.
[0074]
The standard opening command value generation process is performed as follows.
[0075]
First, from the engine temperature Tw detected by the engine temperature sensor 15 at the start of the internal combustion engine 1 (starting mode), the basic value i / ftbl of the standard opening command value θ0 based on a predetermined data table (this is This is the maximum value of the standard opening command value θ0 during operation in the FIRE mode.
[0076]
In this case, in the present embodiment, the basic value i / ftbl is set to the N range (neutral range) when the operation position (hereinafter referred to as the shift position) of the operation lever of the automatic transmission (not shown) of the vehicle during the operation in the FIRE mode. Different values are determined depending on whether or not there is a D range (drive range).
[0077]
That is, when the shift position of the automatic transmission is in the N range, the value ifiret (hereinafter referred to as the N range basic value ifiret) obtained from the engine temperature Tw at the start of the internal combustion engine 1 according to the data table of the solid line a in FIG. Is determined as the basic value i / ftbl of the standard opening command value θ0.
[0078]
Here, the data table of the solid line a is basically set so that the N range basic value ifiret decreases as the engine temperature Tw increases. This is because the engine temperature Tw at the start of the internal combustion engine 1 corresponds to the initial temperature of the catalyst device 3, and the higher the temperature, the more heat is required to perform the desired temperature increase and activation of the catalyst device 3. As a result, the intake air amount of the internal combustion engine 1 can be reduced.
[0079]
When the shift position of the automatic transmission is in the D range, a value iatfire (hereinafter referred to as a D range correction value iatfire) obtained from the engine temperature Tw at the start of the internal combustion engine 1 according to the data table of the solid line b in FIG. ) Is added to the N range basic value ifiret as a basic value i / ftbl (= ifiret + iatfire) of the standard opening command value θ0.
[0080]
Here, the data table of the solid line b is basically set so that the basic value fitbl in the D range is slightly higher than that in the N range at any engine temperature Tw when the internal combustion engine 1 is started. Yes. This is because the load that absorbs the driving force of the internal combustion engine 1 is larger in the D range than in the N range, and the rotational speed of the internal combustion engine 1 is likely to decrease, and the amount of heat of the exhaust gas is greater than in the N range. Because it will decrease.
[0081]
In the present embodiment, the engine temperature Tw at the start of the internal combustion engine 1 is used as the temperature corresponding to the initial temperature of the catalyst device 3 (the temperature state of the catalyst device 3 at the start of the internal combustion engine 1). The initial temperature of the catalyst device 3 may be directly detected when the engine 1 is started, and the basic value i / ftbl of the standard opening command value θ0 may be determined from the detected temperature in the same manner as described above.
[0082]
In this embodiment, since an automatic transmission is adopted for the vehicle, different basic values i / ftbl are set for the N range and the D range. However, a manually operated transmission is used for the vehicle. In this case, a single basic value i / ftbl is determined in accordance with the engine temperature Tw at the start of the internal combustion engine 1 (or the initial temperature of the catalyst device 3) without making such a distinction. You may make it set similarly.
[0083]
In this way, the basic value i / ftbl of the standard opening command value θ0 is determined. On the other hand, in the standard opening command value generation process, it is further determined in advance from the FIRE elapsed time t / fire as shown in FIG. A correction coefficient km / fire (≦ 1) for correcting (multiplying correction) the basic value i / ftbl according to a data table (time table) is obtained for each control cycle. A value obtained by multiplying the basic value i / ftbl by the correction coefficient km / fire is determined as a standard opening command value θ0 (= i / ftbl · km / fire).
[0084]
Here, in the data table of FIG. 7, at the initial stage of the FIRE mode (t / fire: 0 to t1), the standard opening command value θ0 is set to stabilize the combustion state of the internal combustion engine 1 immediately after starting. The correction coefficient km / fire is gradually increased toward “1” so as to gradually increase toward the basic value i / ftbl. Then, after the standard opening command value θ0 reaches the basic value i / ftbl (after the correction coefficient km / fire reaches “1”), the standard opening command value θ0 reaches the standard value i / ftbl for a predetermined time (t / fire: t1 to t2). The correction coefficient km / fire is set to “1” so that the opening command value θ0 is maintained at the basic value i / ftbl. Thereafter (t / fire: t2 to TFIRELMT), the correction coefficient km / fire is gradually decreased so that the standard opening command value θ0 is gradually decreased. Thus, the standard opening command value θ0 is gradually decreased for the following reason.
[0085]
That is, as the warm-up of the internal combustion engine 1 proceeds to some extent, the friction (friction) of each part of the internal combustion engine 1 gradually decreases, and the increasing tendency of the rotational speed Ne of the internal combustion engine 1 increases. As a result, the ignition The ignition timing of the internal combustion engine 1 operated by the timing operation rotational speed F / B control is shifted to the retard side. At this time, if the ignition timing of the internal combustion engine 1 reaches the limit value on the retarded side where the ignition timing can be operated while operating the internal combustion engine 1 normally, the rotational speed Ne of the internal combustion engine 1 It becomes impossible to suppress the upward trend. In order to prevent such a situation, after the FIRE elapsed time t / fire reaches the predetermined time t2 and the operation of the FIRE mode has progressed to some extent (after the warm-up of the internal combustion engine 1 has progressed to some extent), the standard opening degree The command value θ0 (intake air amount of the internal combustion engine 1) is gradually decreased to suppress the increasing tendency of the rotational speed Ne due to the decrease in friction.
[0086]
The contents described above are the contents of the standard opening command value generation process.
[0087]
In the system of the present embodiment, basically, by operating the bypass opening according to the standard opening command value θ0 generated in a feed-forward manner as described above, the operation of the internal combustion engine 1 is stably performed, It is possible to appropriately perform desired temperature increase and activation of the catalyst device 3.
[0088]
Next, the intake air amount F / B control correction process and the atmospheric condition correction process will be described.
[0089]
In the present embodiment, the standard opening command value θ 0 determined as described above in the standard opening command value generation process is the actual opening degree of the bypass valve 7 with respect to the bypass opening command value given to the bypass valve actuator 24. The actual intake air amount in the combustion chamber 4 of the internal combustion engine 1 is uniquely determined by a standard fixed correlation, and the atmospheric pressure Pa and the atmospheric temperature Ta are each a predetermined standard atmospheric pressure (for example, 1 atmospheric pressure). The standard air temperature (for example, a room temperature of 25 ° C.) is determined on the premise of an ideal condition.
[0090]
However, the actual opening of the bypass valve 7 or the actual intake air amount with respect to the command value of the bypass opening is caused by variations in the operating characteristics of the bypass valve actuator 24 and the bypass valve 7 or changes in characteristics over time. Variation occurs (hereinafter, such variation in intake air amount is referred to as variation due to structural factors).
[0091]
Even if there is no variation due to the structural factors as described above, the actual intake air amount with respect to the command value of the bypass opening varies depending on the atmospheric pressure Pa. Furthermore, although it is smaller than the influence of the atmospheric pressure Pa, the actual intake air amount with respect to the command value θCMD of the bypass opening varies depending on the atmospheric temperature Ta (hereinafter, depending on the atmospheric pressure Pa and the atmospheric temperature Ta). Variation in intake air volume is called variation due to atmospheric conditions).
[0092]
That is, when the bypass opening command value is constant (the bypass opening is constant), the actual intake air amount decreases as the atmospheric pressure Pa decreases. Further, the higher the atmospheric temperature Ta, the smaller the air density. Therefore, the actual intake air amount (intake air mass) decreases as the atmospheric temperature Ta increases.
[0093]
When such a variation in the intake air amount occurs, a variation in the amount of heat of the exhaust gas generated by the internal combustion engine 1 (which is approximately proportional to the intake air amount) also occurs, so that a variation in the temperature raising mode of the catalyst device 3 also occurs. For this reason, in some cases, it is not possible to reliably raise the temperature and activate the catalyst device 3 which is the main purpose of increasing the amount of intake air in the FIRE mode, so that the catalyst device 3 is in operation during the operation of the FIRE mode. There is a risk that the required purification performance cannot be ensured.
[0094]
The intake air amount F / B control correction process and the atmospheric condition correction process are processes performed to eliminate such inconveniences. Of these, the intake air amount F / B control correction process is subject to variations due to the above-described structural factors. The process for compensation and the atmospheric condition correction process are processes for compensating for variations due to atmospheric conditions.
[0095]
First, an intake air amount F / B control correction process for compensating for variations due to the structural factors will be described.
[0096]
As a basic idea in the intake air amount F / B control correction processing, heat amount data representing the amount of heat actually given to the catalyst device 3 by the exhaust gas of the internal combustion engine 1 is sequentially (during each control cycle) during the operation of the FIRE mode. Detected or estimated and acquired, and the command value of the bypass opening is corrected by feedback control processing so that the value of the heat quantity data converges to a predetermined target value (this corresponds to the target heat quantity to be given to the catalyst device 3) To do. Then, by operating the bypass opening based on the corrected command value, the amount of heat actually given to the catalyst device 3 is matched with the target amount of heat corresponding to the target value of the value of the heat amount data representing it, thereby The variation in the temperature rising mode of the catalyst device 3 is eliminated.
[0097]
In this case, the heat amount data representing the amount of heat given to the catalyst device 3 is, for example, an intake air amount or fuel supply amount for each control cycle (instantaneous instant) (these are basically (for each control cycle given to the catalyst device 3) Proportional to the amount of heat (instantaneous instantaneous)), or their integrated value (which is proportional to the integrated value of instantaneous instantaneous amount of heat applied to the catalytic device 3), or the temperature rise of the catalytic device 3 (initial temperature of the catalytic device 3) For example, the amount of temperature rise from is proportional to the integrated value of instantaneous instantaneous heat amount applied to the catalyst device 3).
[0098]
In this embodiment, for example, an integrated value of the intake air amount is used as the heat amount data, and the integrated value of the actual intake air amount of the internal combustion engine 1 during the operation of the FIRE mode is as follows for each control cycle. The target value of the estimated value is set for each control cycle as follows.
[0099]
First, regarding the estimation of the integrated value of the intake air amount, the intake air amount sucked into the combustion chamber 4 of the internal combustion engine 1 per 1 TDC, which is the control cycle in the present embodiment, is approximately that of the chamber 13 shown in FIG. It is proportional to the internal pressure, that is, the intake pressure Pb.
[0100]
Therefore, in this embodiment, the estimated value gair / pre of the intake air amount for each control cycle is obtained by the following equation (1) (hereinafter, this estimated value gair / pre is referred to as the estimated intake air amount gair / pre).
[0101]
[Expression 1]

[0102]
Here, the coefficient Ga1 in the equation (1) is a predetermined value (a constant value) determined in advance.
[0103]
Then, the estimated intake air amount gair / pre is cumulatively added for each control cycle according to the following equation (2) during the operation of the FIRE mode, thereby obtaining an integrated value qair / pre of the intake air amount (hereinafter, This integrated value qair / pre is referred to as an estimated integrated intake air amount qair / pre).
[0104]
[Expression 2]

[0105]
Here, k in the formula (2) is the number of control cycles (hereinafter the same).
[0106]
The integrated value of the actual intake air amount of the internal combustion engine 1 may be obtained, for example, by directly detecting the intake air amount for each control cycle using an air flow sensor or the like and integrating the intake air amount.
[0107]
Next, the target value of the integrated value of the intake air amount (hereinafter referred to as the target integrated intake air amount qair / cmd) corresponding to the target value of the integrated value of the heat amount given to the catalyst device 3 is the temperature rise / activity of the catalyst device 3. In order to properly perform the conversion, various forms can be set. However, the target value affects the intake air amount of the internal combustion engine 1 during the operation of the FIRE mode, and thus the combustion state, emission state, etc. of the internal combustion engine 1. Need to be considered.
[0108]
Therefore, in the present embodiment, the standard opening determined so that the temperature rise and activation of the catalyst device 3 can be appropriately performed and the operation of the internal combustion engine 1 can be stably performed under the ideal conditions as described above. The target integrated intake air amount qair / cmd is set based on the degree command value θ0.
[0109]
That is, the standard opening command value θ0 is a standard atmospheric pressure and an atmospheric pressure Pa and an atmospheric temperature Ta that are constant, with the actual bypass opening and intake air amount being uniquely determined with respect to the bypass opening command value. Under ideal conditions such as a standard atmospheric temperature, the temperature and activation of the catalyst device 3 are appropriately performed and the operation of the internal combustion engine 1 can be performed stably. In other words, the standard opening command value θ0 is the optimum intake air to be sucked into the combustion chamber 4 in order to properly raise the temperature and activate the catalyst device 3 and to operate the internal combustion engine 1 stably. The amount is specified.
[0110]
Therefore, the intake air amount for each control cycle of the combustion chamber 4 of the internal combustion engine 1 when the bypass opening is operated in accordance with the standard opening command value θ0 under the ideal conditions as described above is set as the target for each control cycle. The intake air amount gair / cmd may be set, and the integrated value of the target intake air amount gair / cmd may be set as the target integrated intake air amount qair / cmd.
[0111]
In this case, the target intake air amount gair / cmd and the target integrated intake air amount qair / cmd can be obtained from the standard opening command value θ0 as follows for each control cycle.
[0112]
First, when the actual bypass opening is θ, the air amount Gi per unit time (fixed time) passing through the bypass valve 7 is the atmospheric pressure Pa that is the pressure on the upstream side of the bypass valve 7 and the pressure on the downstream side. Is generally expressed by the following equation (3) using the intake pressure Pb.
[0113]
[Equation 3]

[0114]
Here, Ci in the formula (3) is a coefficient corresponding to the air density, and the air density corresponds to the atmospheric temperature Ta. Strictly speaking, the term θ in the expression (3) is the effective opening area of the bypass valve 7, but here, the bypass opening θ is used instead of the effective opening area. Note that the correction of the influence by using the bypass opening θ instead of the effective opening area may be included in the coefficient Ci.
[0115]
In this case, when the bypass opening is operated according to the standard opening command value θ0 under the ideal conditions, θ = θ0, Pa = standard atmospheric pressure (constant) in the equation (3), and the coefficient Ci Basically, it becomes a constant value according to the standard atmospheric temperature. Further, in the steady operation state of the internal combustion engine 1 in the FIRE mode, the change in the intake pressure Pb is relatively small and becomes a substantially constant value. Further, in the steady operation state of the internal combustion engine 1 in the FIRE mode, the throttle valve 5 is basically closed, and the intake air amount of the combustion chamber 4 of the internal combustion engine 1 is the air passing through the bypass valve 7. It can be regarded as being equal to the quantity Gi.
[0116]
Therefore, the intake air amount per unit time (constant time) of the combustion chamber 4 when the bypass opening is operated according to the standard opening command value θ0 under ideal conditions is proportional to the standard opening command value θ0.
[0117]
Therefore, under ideal conditions, the intake air amount per control cycle (1 TDC) of the combustion chamber 4 when the bypass opening is operated according to the standard opening command value θ0, that is, the target intake air amount gair / cmd is The following equation (4) can be obtained.
[0118]
[Expression 4]

[0119]
Here, the expression (4) includes the term of the reciprocal number (1 / Ne) of the rotational speed Ne of the internal combustion engine 1 in this embodiment, the time of one control cycle (1 TDC) is inversely proportional to the rotational speed Ne. Because. Further, as is clear from the above, the parameter Ga2 in the equation (4) is the standard atmospheric pressure, the standard atmospheric temperature, the standard intake pressure Pb in the steady operation state of the internal combustion engine 1 in the FIRE mode, and the like. Is a predetermined constant determined by.
[0120]
In the present embodiment, the target integrated air amount gair / cmd obtained by the equation (4) is cumulatively added for each control cycle by the following equation (5) during the operation of the FIRE mode, whereby the target integrated air amount is calculated. Obtain the intake air volume qair / cmd.
[0121]
[Equation 5]

[0122]
Since the target integrated intake air amount qair / cmd obtained in this way is determined by the standard opening command value θ0, the engine temperature Tw at the start of the internal combustion engine 1 (or the initial temperature of the catalyst device 3). Depending on. Corresponding to the target integrated intake air amount qair / cmd, the intake air amount for each control cycle to be sucked into the combustion chamber 4 of the internal combustion engine 1, that is, the target intake air amount gair / cmd is a standard opening degree. It changes in accordance with the FIRE elapsed time t / fire in the form of temporal change similar to the command value θ0.
[0123]
Further, the target integrated intake air amount qair / cmd as described above may be set in advance by a time table, and may be obtained from the FIRE elapsed time t / fire for each control cycle using this time table.
[0124]
In the system of the present embodiment, the intake air amount F / B control correction process basically includes the estimated integrated intake air amount qair / pre obtained as described above (this is the amount of heat actually applied to the catalyst device 3). (Corresponding to the integrated value) to converge to the target integrated intake air amount qair / cmd (which corresponds to the target value of the integrated value of the amount of heat to be applied to the catalyst device 3) (so as to eliminate these deviations) By calculating the correction value of the bypass opening command value by feedback control processing, and correcting the standard opening command value θ0 by this correction amount to determine the bypass opening command value, the intake air amount due to the structural factors described above Is compensated for, and as a result, the variation in the temperature rise mode of the catalyst device 3 is eliminated.
[0125]
In this case, in the intake air amount F / B control correction process, as will be described later, the rate of decay of the deviation between the estimated integrated intake air amount qair / pre and the target integrated intake air amount qair / cmd (the estimated integrated intake air amount qair / pre Since it is preferable to variably set the target integrated intake air amount qair / cmd) according to the situation, the sliding mode as a response designation type control that can arbitrarily designate the decay speed as a desired decay speed. Control (more specifically, adaptive sliding mode control) is used for the feedback control process.
[0126]
By using this adaptive sliding mode control (hereinafter referred to as intake side adaptive SLD control), the command value of the bypass opening is corrected so that the estimated integrated intake air amount qair / pre converges to the target integrated intake air amount qair / cmd. The algorithm of the intake air amount F / B control correction process is constructed as follows in the present embodiment. In the following description, for convenience of explanation, the integrated value of the actual intake air amount of the combustion chamber 4 of the internal combustion engine 1 including the estimated integrated intake air amount qair / pre is used for a while using the reference sign Qa. The integrated intake air amount Qa is referred to, and the target value is referred to as a target integrated amount q (which corresponds to the target integrated intake air amount qair / cmd) using the reference symbol q. Further, the command value of the bypass opening is generally referred to as an opening command Θ using the reference sign Θ.
[0127]
First, the actual intake air amount for each control cycle of the combustion chamber 4 of the internal combustion engine 1 is expressed using the reference symbol Gcyl, and the correlation between the intake air amount Gcyl and the opening degree command Θ is expressed by the following equation (6). As described above, it is expressed by a discrete system (discrete time system) model (first order autoregressive model) of a first order lag system.
[0128]
[Formula 6]

[0129]
Here, α and β in the equation (6) are model parameters corresponding to the atmospheric pressure Pa, the atmospheric temperature Ta, the intake pressure Pb, the rotational speed Ne, and the like.
[0130]
Further, the integrated intake air amount Qa for each control cycle is expressed by the following equation (7).
[0131]
[Expression 7]

[0132]
The following equation (8) is obtained by using this equation (7) and the above equation (6).
[0133]
[Equation 8]

[0134]
Here, Gcyl (k) in the equation (8) is Gcyl (k) = Qa (k) −Qa (k−1) according to the equation (7), so this is substituted into the equation (8). Then, the following formula (9) is obtained.
[0135]
[Equation 9]

[0136]
This equation (9) is a discrete system model (the system that generates the integrated intake air amount Qa from the opening degree command Θ, that is, the system to be controlled by the intake side adaptive SLD control in the system of the present embodiment). A second-order autoregressive model (hereinafter referred to as an intake control target model). That is, the intake control target model uses the integrated intake air amount Qa (k + 1) in each control cycle as the output of the control object of the intake side adaptive SLD control, and the integrated intake air amount Qa in the past control cycle. This is expressed using time series data Qa (k), Qa (k-1) and an opening degree command Θ (k) as an input to be controlled by the intake side adaptive SLD control. In equation (9), the coefficients a1 and a2 related to the cumulative intake air amounts Qa (k) and Qa (k-1) and the coefficient b1 related to the opening degree command Θ (k) It is a model parameter that defines behavior characteristics.
[0137]
In the present embodiment, an algorithm for intake-side adaptive SLD control is constructed as follows based on the intake control target model.
[0138]
That is, in the intake-side adaptive SLD control, the switching function σ1 required for the sliding mode control is obtained by using time series data Eq () for each control cycle of the deviation Eq = Qa−q between the integrated intake air amount Qa and the target integrated amount q. It is defined by the linear function of the following equation (10) with k) and Eq (k-1) as variables.
[0139]
[Expression 10]

[0140]
Here, s1 and s2 in the equation (10) are coefficient parameters of each term of the switching function σ1, and are set so as to satisfy the condition of the following equation (11).
[0141]
## EQU11 ##

[0142]
In the present embodiment, s1 = 1 is set for simplification. In this embodiment, the value of the coefficient parameter s2 (more generally, the value of s2 / s1) is variably set. This will be described later.
[0143]
When the switching function σ1 is defined in this way, state quantities (Eq (k), Eq (k-1)) consisting of a set of time series data Eq (k) and Eq (k-1) of the deviation Eq are shown in FIG. As shown in FIG. 8, when convergence is made on a switching line defined by the relational expression σ 1 = 0 (this is also called a slip line) and the convergence state is maintained, state quantities (Eq (k), Eq (k− 1)) can be converged very stably to the equilibrium point on the switching line σ1 = 0, that is, the point where Eq (k) = Eq (k-1) = 0, regardless of the influence of disturbance or the like. .
[0144]
In this embodiment, since the phase space related to the switching function σ1 is two-dimensional (two components of the state quantities (Eq (k), Eq (k-1))), the switching line σ1 = 0 is a straight line However, when the phase space is three-dimensional, the switching line is a plane, and in this case, the switching line is sometimes referred to as a slip plane. Further, when the phase space becomes higher than four dimensions, the switching line becomes a hyperplane that cannot be geometrically illustrated.
[0145]
On the other hand, in order to converge the integrated intake air amount Qa to the target integrated amount q, a control input generated by the intake side adaptive SLD control as an input to be given to the control target modeled by the equation (9), that is, the opening degree The command Θ is basically an equivalent control input determined based on a control law for constraining the state quantity (Eq (k), Eq (k-1)) to the switching line σ1 = 0 determined as described above. Θeq, a reaching law input Θrch determined based on a reaching law that is a control law for converging the state quantities (Eq (k), Eq (k-1)) to the switching line σ 1 = 0, and the state quantity (Eq ( k), Eq (k-1)) and the sum of the adaptive law input Θadp determined based on the adaptive law, which is a control law for eliminating the influence of disturbances and the like when converging to the switching line σ1 = 0. (See the following equation (12)).
[0146]
[Expression 12]

[0147]
In normal sliding mode control, the adaptive law is not considered, and in this case, the adaptive law input Θadp is omitted.
[0148]
In this case, the equivalent control input Θeq is given by the following equation (13).
[0149]
[Formula 13]

[0150]
This equation (13) is a condition that σ1 (k) = σ1 (k-1) for constraining the state quantity (Eq (k), Eq (k-1)) to the switching line σ1 = 0. This can be derived based on the equation (9) of the intake control target model.
[0151]
In addition, the reaching law input Θrch and the adaptive law input Θadp can be determined by various methods. It is assumed that the value of the switching function σ1 is proportional to the integrated value (integrated value), and is determined by the following equations (14) and (15), respectively.
[0152]
[Expression 14]

[0153]
[Expression 15]

[0154]
Here, F1 in the equation (14) is a coefficient that defines a gain related to the reaching law, and may be set so as to satisfy the condition of the following equation (16). However, in order to reduce chattering when the value of the switching function σ1 converges to the switching line σ1 = 0, it is preferable to set the coefficient F1 so as to satisfy the condition of the equation (16) ′.
[0155]
[Expression 16]

[0156]
Further, F2 in the equation (15) is a coefficient that defines a gain related to the adaptive law, and may be set so as to satisfy the condition of the following equation (17). In the equation (17), ΔT is a control cycle (control cycle).
[0157]
[Expression 17]

[0158]
In the algorithm of the intake side adaptive SLD control used in the present embodiment, basically, the equivalent control input Θeq, the reaching law input Θrch and the adaptive law input Θadp are obtained by the equations (13) to (15), and the sum of them is calculated. By generating the degree command Θ, it is possible to manipulate the intake air amount during the operation of the FIRE mode so that the integrated intake air amount Qa converges to the target integrated amount q.
[0159]
By the way, in order to obtain the equivalent control input Θeq, the reaching law input Θrch and the adaptive law input Θadp by the equations (13) to (15), respectively, the model parameters a1,. It is necessary to identify the values of a2 and b1. However, the values of these model parameters a1, a2, and b1 are affected by various factors during the operation of the FIRE mode, and it is difficult to optimally identify these values.
[0160]
For this reason, in the present embodiment, the model parameters a1, a2, and b1 are eliminated as follows, and a simplified algorithm for intake-side adaptive SLD control is constructed.
[0161]
First, regarding the reaching law input Θrch and the adaptive law input Θadp, the model parameter included in the above equations (14) and (15) for obtaining them is only b1. Then, if (F1 / b1) is replaced with Fx in equation (14), and (F2 / b1) is replaced with Fy in equation (15), equations (14) and (15) are respectively 18) and (19).
[0162]
[Formula 18]

[0163]
[Equation 19]

[0164]
Therefore, the reaching law input Θrch and the adaptive law input Θadp can be obtained from these equations (18) and (19) without using the model parameter b1.
[0165]
In this case, the values of the coefficients Fx and Fy in the equations (18) and (19) are measured in consideration of the stability of convergence of the value of the switching function σ1 to the switching line σ1 = 0 and the rapid response. Or through simulation.
[0166]
Next, regarding the equivalent control input Θeq, the equation (13) for obtaining this can be rewritten into the following equation (20) using the deviation Eq = Qa−q.
[0167]
[Expression 20]

[0168]
The term including the first square bracket in the equation (20) is a feedback term based on the deviation Eq between the cumulative intake air amount Qa and the target cumulative amount q, and the term including the second square bracket is the target This is a feed-forward term based only on the integrated amount q. Hereinafter, the feedback term and the feedforward term are set as Θeq / fb and Θeq / ff as shown in the following equations (21) and (22), respectively.
[0169]
[Expression 21]

[0170]
[Expression 22]

[0171]
Here, the feedforward term Θeq / ff is an input (opening degree command Θ) that should be given to the controlled object in a state where the deviation Eq is constantly “0”. On the other hand, the standard opening command value θ0 defines the target integrated amount q in the present embodiment, and the intake air amount and thus the integrated intake air amount with respect to the standard opening command value θ0 is the target. It is set in a feedforward manner as it is uniquely determined by the value to be obtained.
[0172]
Therefore, the feedforward term Θeq / ff in equation (20) can be replaced with a standard opening command value θ0 that does not include the model parameters a1, a2, and b1.
[0173]
Further, regarding the feedback term Θeq / fb, the equivalent control input Θeq including the feedback term Θeq / fb causes the state quantities (Eq (k), Eq (k-1)) to be placed on the switching line σ1 = 0. Although it is a control input for restraining, according to various studies by the inventors of the present application, in the system of this embodiment, the state quantities (Eq (k), Eq (k-1)) are the switching line σ1 = 0. High stability in the state existing in the vicinity. Further, in the present embodiment, by adopting the adaptive law input Θadp, the stability of convergence of the state quantities (Eq (k), Eq (k-1)) on the switching line σ1 = 0 can be improved. it can.
[0174]
For this reason, in the system of this embodiment, even if the feedback term Θeq / fb is omitted, it is considered that the controllability is not impaired in practice.
[0175]
Considering this, it is considered that the equivalent control input Θeq may omit the feedback term Θeq / fb and replace the feedforward term Θeq / ff with the standard opening command value θ0. In this case, the equivalent control input Θeq can be obtained without using the model parameters a1, a2, and b1.
[0176]
Therefore, in this embodiment, the equivalent control input Θeq of the intake side adaptive SLD control is given by the following equation (23).
[0177]
[Expression 23]

[0178]
From the above, in the intake air amount F / B control correction processing in the present embodiment, the input to the control object determined by the intake side adaptive SLD control in order to compensate for the variation in the intake air amount due to the structural factors described above, that is, The opening degree command Θ is obtained by calculation of the following equation (24).
[0179]
[Expression 24]

[0180]
In other words, in this embodiment, the sum (= Θrch + Θadp) of the reaching law input Θrch and the adaptive law input Θadp obtained by the equations (18) and (19) is bypassed as shown in the following equation (25). Command value correction amount i / sld (hereinafter referred to as SLD opening correction amount i / sld), and the standard opening command value θ0 is corrected by this SLD opening correction amount i / sld (standard opening command value). By adding the SLD opening correction amount i / sld to the value θ0), an opening command Θ for compensating for variations due to structural factors is obtained.
[0181]
[Expression 25]

[0182]
In this case, in the present embodiment, the value of the switching function σ1 required for obtaining the SLD opening correction amount i / sld (= Θrch + Θadp) is the above-mentioned equation as the integrated intake air amount Qa of the equation (10). Using the estimated integrated intake air amount qair / pre obtained by (2) and calculating the following equation (26) using the target integrated intake air amount qair / cmd obtained by the above equation (5) as the target integrated amount q. .
[0183]
[Equation 26]

[0184]
In the present embodiment, the calculation of the SLD opening correction amount i / sld is interrupted (SLD open) while the FIRE mode is interrupted (the FIRE interrupt flag f / fpause is set to “1”). However, the estimated integrated intake air amount qair / pre and the target integrated intake air amount qair / cmd will continue to be calculated (however, the estimated integrated intake air will be maintained). The quantity qair / pre is excluded during the fuel cut of the internal combustion engine 1). If the interruption operation is canceled before the FIRE time limit TFIRELMT elapses, the calculation of the SLD opening correction amount i / sld and the correction of the standard opening command value θ0 corresponding thereto are resumed. To do.
[0185]
This is done for the following reason. That is, during the interruption operation of the FIRE mode, basically, the vehicle is traveling or overlaid by the internal combustion engine 1 by operating the accelerator pedal of the vehicle. In this situation, since the throttle valve 5 is opened to an opening corresponding to the accelerator operation amount Ap, the actual intake air amount in the combustion chamber 4 of the internal combustion engine 1 becomes the intake air amount via the bypass valve 7. The intake air amount through the throttle valve 5 is added.
[0186]
In such a situation, the estimated integrated intake air amount qair / pre is an integrated value of the intake air amount by both the bypass valve 7 and the throttle valve 5, and is thus set to the standard opening command value θ 0 of the bypass valve 7. It is not preferable to operate the bypass opening so as to converge to the target integrated intake air amount qair / cmd determined accordingly in order to ensure the operation performance of the internal combustion engine 1 according to the operation of the accelerator pedal. For this reason, in the present embodiment, the calculation of the SLD opening correction amount i / sld is interrupted during the interruption operation in the FIRE mode.
[0187]
Further, during the FIRE mode interruption operation, the actual intake air amount in the combustion chamber 4 of the internal combustion engine 1 is obtained by adding the intake air amount via the throttle valve 5 to the intake air amount via the bypass valve 7. Therefore, the amount of heat given to the catalyst device 3 is further increased. Therefore, the catalyst device 3 may be sufficiently heated and activated during the interruption operation in the FIRE mode. However, in many cases, the interruption operation is canceled in a short time. In some cases, the temperature rise and activation of the apparatus 3 may still be insufficient. For this reason, in the present embodiment, the estimated integrated intake air amount qair / pre and the target integrated intake air amount qair / cmd are continuously calculated even during the interruption operation in the FIRE mode, and the SLD is canceled after the interruption operation is canceled. By restarting the calculation of the opening correction amount i / sld and correcting the standard opening command value θ 0, the temperature increase and activation of the catalyst device 3 during the operation of the FIRE mode is ensured. However, in this case, regarding the estimated integrated intake air amount qair / pre, during the fuel cut of the internal combustion engine 1, the air sucked into the combustion chamber 4 does not contribute to the amount of heat given to the catalyst device 3. Does not calculate the estimated integrated intake air amount qair / pre.
[0188]
Further, in the present embodiment, the calculation of the SLD opening correction amount i / sld is interrupted even during the ignition timing response correction process described later (the value of the SLD opening correction amount i / sld is set to the value of the interruption operation). Keep the previous value). This is due to the following reason. That is, the ignition timing response correction process, which will be described in detail later, is a process of correcting the opening degree command Θ to the decrease side in a feed-forward manner with respect to the standard opening degree command value θ0. In this case, if the SLD opening correction amount i / sld is calculated, the SLD opening correction amount i / sld is calculated so as to cancel out the decrease in the opening command Θ by the ignition timing response correction processing. is there.
[0189]
The contents described above are the basic contents of the intake air amount F / B control correction process.
[0190]
By the way, in the present embodiment, the stage immediately after starting the increase of the intake air amount of the internal combustion engine 1, that is, the stage of increasing the bypass opening is a state immediately after the start of the internal combustion engine 1, and thus If the standard opening command value θ0 is largely corrected by the SLD opening correction amount i / sld at any stage, the combustion state of the air-fuel mixture in the combustion chamber 4 of the internal combustion engine 1 and the emission state of the internal combustion engine 1 are deteriorated. There is a fear. Further, in the present embodiment, since the target integrated intake air amount qair / cmd is based on the assumption of a steady intake state of the internal combustion engine 1, the target integrated intake air amount qair is immediately after the start of the increase of the intake air amount. I think the reliability of / cmd is poor. For this reason, immediately after the start of the increase of the intake air amount, the deviation Eq between the target integrated intake air amount qair / cmd and the estimated integrated intake air amount qair / pre becomes large, and the SLD opening correction amount i / sld may also be large.
[0191]
In consideration of this, in the intake air amount F / B control correction process in the present embodiment, after the start of the increase of the intake air amount (after the start of the FIRE mode), it is referred to as a predetermined time TISLDLMT (SLD correction limit time TISLDLMT). 7) (until FIRE elapsed time t / fire ≧ TISLDLMT), the target integrated intake air amount qair / cmd and the estimated integrated intake air amount qair / pre are forcibly set to “0”. (Eq = 0 at this time). In this way, in the state immediately after the start of the internal combustion engine 1 until the FIRE elapsed time t / fire reaches the SLD correction limit time TISLDLMT, the value of the SLD opening correction amount i / sld is also maintained at “0”. The standard opening command value θ0 is not corrected by the SLD opening correction amount i / sld.
[0192]
Further, in the intake air amount F / B control correction process in the present embodiment, if the correction is performed rapidly even immediately after starting the actual correction of the standard opening command value θ0 with the SLD opening correction amount i / sld, There is a possibility that the combustion state and emission performance of the internal combustion engine 1 may be impaired. For this reason, in this embodiment, using the response designation characteristic of the intake side adaptive SLD control used in the intake air amount F / B control correction process, the estimated integrated intake air amount qair / pre and the target integrated intake air amount qair / cmd The rate of decay of the deviation Eq (hereinafter referred to as intake deviation Eq) (the convergence speed of the estimated integrated intake air amount qair / pre to the target integrated intake air amount qair / cmd) is variably set as follows. ing.
[0193]
That is, in the intake-side adaptive SLD control, when the state quantity (Eq (k), Eq (k-1)) related to the intake deviation Eq converges to the switching line σ1 = 0, it is clear from the equation (10). Thus, the following relational expression (27) is established.
[0194]
[Expression 27]

[0195]
Therefore, the value of the ratio (s2 / s1) of the coefficient parameters s1, s2 of the switching function σ1 (where -1 <s2 / s1 <1) defines the rate of decay of the intake deviation Eq to "0". (As || s2 / s1 | approaches "0", the decay rate increases.) Therefore, the decay rate of the intake air deviation Eq can be designated by the value of the ratio (s2 / s1), which is the response designation characteristic of the intake side adaptive SLD control.
[0196]
Note that increasing the decay rate of the intake deviation Eq is equivalent to increasing the gain of feedback control by the intake-side adaptive SLD control. Equation (27) expresses a first-order lag system without input, and the ratio (s 2 / s 1) corresponds to the pole of the first-order lag system (hereinafter, the ratio (s 2 / s 1)). ) Is called pole pole / i). In this embodiment, s1 = 1 is set, and in this case, pole / i = s2. The attenuation of the intake air deviation Eq to “0” is preferably non-oscillating. Therefore, in this embodiment, s 2 / s 1 = pole / i <0 (s 2 / s 1> 0). As is clear from the equation (27), the attenuation of the intake air deviation Eq to “0” becomes oscillating).
[0197]
In the intake air amount F / B control correction process in the present embodiment, such a response designating characteristic of the intake side adaptive SLD control is used. Basically, from the FIRE elapsed time t / fire, as shown in FIG. By setting a value pole / itbl (hereinafter referred to as a pole table value pole / itbl) obtained for each control cycle based on a defined data table (time table) as the value of the pole pole / i, the pole pole / i Is variably set according to the FIRE elapsed time t / fire.
[0198]
Here, in the data table of FIG. 9, after the FIRE elapsed time t / fire reaches a predetermined value TPOLEVST (however, TPOLEVST ≧ SLD correction limit time TISLDLMT), the pole table value pole / itbl and eventually the pole pole / i are changed to FIRE. As the elapsed time t / fire increases, the predetermined lower limit value pole / i0 (<0. In this embodiment, “−1”) is changed to the predetermined steady value pole / ix (pole / i0 <pole / ix <0). (The absolute value of the pole table value pole / itbl is gradually reduced) and after reaching the steady value pole / ix (the FIRE elapsed time t / fire becomes the predetermined value TPOLEX in FIG. 9) After reaching this value, the steady value pole / ix is maintained. Thus, in this embodiment, immediately after the correction of the standard opening command value θ0 by the SLD opening correction amount i / sld is started, the decay rate of the intake deviation Eq is gradually increased (estimated integrated intake air amount qair / pre, the target integrated intake air amount qair / cmd is gradually converged), and until the FIRE elapsed time t / fire reaches the predetermined value TPOLEX than after the predetermined value TPOLEX is reached. The decay rate of the intake air deviation Eq is slowed down.
[0199]
The increase of the pole pole / i by the pole table value pole / itbl is basically performed in the initial stage after starting the internal combustion engine 1 (the stage of increasing the intake air amount).
[0200]
In the present embodiment, the intake air amount F / B control correction process (more precisely, the calculation process of the SLD opening correction amount i / sld) is not performed during the interruption operation of the FIRE mode. In the state, the pole pole / i is set to the lower limit value pole / i0 of the pole table value pole / itbl. When the interruption operation of the FIRE mode is released, the pole pole / i is gradually returned from the lower limit value pole / i0 toward the pole table value pole / itbl (see the virtual line in FIG. 9). Yes.
[0201]
Next, the atmospheric condition correction process for compensating for the influence of the variation in the intake air amount due to the atmospheric conditions described above will be described.
[0202]
First, the atmospheric pressure in the atmospheric conditions will be described. In this description, it is assumed that the atmospheric temperature is constant and the bypass opening is equal to the opening command Θ.
[0203]
If the pressure in the combustion chamber 4 of the internal combustion engine 1 is Pcyl and the opening (effective opening area) of the intake valve 11 (see FIG. 2) is Acyl, the intake air amount Gcyl of the combustion chamber 4 is equal to these pressures Pcyl. Generally, it is expressed by the following equation (28) using the opening degree Acyl and the intake pressure Pb.
[0204]
[Expression 28]

[0205]
Here, Ci is a coefficient corresponding to the air density, as described with respect to the equation (3).
[0206]
In this case, in the internal combustion engine 1, the pressure in the combustion chamber 4 in the equation (28) is Pcyl, and the opening degree Acyl of the intake valve 11 is basically constant. Further, the coefficient Ci may be considered constant if the atmospheric temperature Ta is constant.
[0207]
Therefore, it can be seen from the equation (28) that the intake pressure Pb needs not to change according to the atmospheric pressure Pa in order to make the intake air amount Gcyl in the combustion chamber 4 independent of the atmospheric pressure. .
[0208]
On the other hand, the air amount Gi that passes through the bypass valve 7 with respect to the opening degree command Θ of the bypass opening degree is expressed by the following equation (29) as in the above equation (3).
[0209]
[Expression 29]

[0210]
Further, in a state where the atmospheric pressure Pa is a predetermined standard atmospheric pressure (hereinafter referred to as reference numeral Pa0), the opening degree command Θ is set to the standard opening degree command value θ0 (this is the standard atmospheric pressure Pa0 as described above). If the air amount passing through the bypass valve 7 is Gi0 (this corresponds to the target intake air amount gair / cmd), the air amount Gi0 (hereinafter referred to as standard air) is assumed. The quantity Gi0) is expressed by the following equation (30).
[0211]
[30]

[0212]
Here, Pb0 in the equation (30) is an intake pressure (hereinafter referred to as standard intake pressure) generated in the chamber 13 (see FIG. 2) when the bypass opening is set to the standard opening command value θ0 at the standard atmospheric pressure Pa0. Pressure Pb0).
[0213]
Further, according to the well-known gas equation of state, the condition for the intake pressure Pb in the chamber 13 not to fluctuate is that the amount of air flowing into the chamber 13, that is, the amount of air Gi passing through the bypass valve 7 and the chamber 13 The amount of air flowing out, that is, the amount of intake air Gcyl in the combustion chamber 4 becomes equal.
[0214]
From the above, when the atmospheric pressure Pa fluctuates with respect to the standard atmospheric pressure Pa0, the opening degree command Θ for preventing the intake air amount Gcyl of the combustion chamber 4 from changing is the intake pressure in the equation (29). What is necessary is just to determine that Pb becomes equal to the standard intake pressure Pb0, and the air amount Gi represented by the equation (29) becomes equal to the standard air amount Gi0 (equation (30)).
[0215]
That is, the opening degree command Θ may be determined so as to satisfy the condition of the following expression (31).
[0216]
[31]

[0217]
When the equation (31) is solved for the opening degree command Θ, the following equation (32) is obtained.
[0218]
[Expression 32]

[0219]
Therefore, basically, if the opening degree command Θ is determined by correcting the standard opening degree command value θ0 based on the equation (32), it is possible to compensate for variations in the intake air amount due to atmospheric pressure. In other words, the opening degree for compensating the variation in the intake air amount due to the atmospheric pressure by multiplying the opening degree command by the square root value kpa (hereinafter referred to as the atmospheric pressure correction coefficient kpa) in the equation (32) and correcting it. Directives can be determined.
[0220]
By the way, in this embodiment, since the standard opening command value θ0 is temporally changed as described above, the standard intake pressure Pb0 used in the calculation of the equation (32) also changes, and accordingly, the intake air amount due to the atmospheric pressure is changed. In order to accurately compensate the variation, it is considered preferable to appropriately change the value of the standard intake pressure Pb0 according to the standard opening command value θ0 using a predetermined data table or the like. However, in the present embodiment, during the steady operation of the internal combustion engine 1 in the FIRE mode, the fact that the fluctuation of the intake pressure Pb is practically small and the stability of the control system is taken into consideration, the equation (32) A predetermined value (fixed value) determined in advance is used as the standard intake pressure Pb0.
[0221]
Further, in the present embodiment, in order to reduce the calculation load of the controller 2, actually, the following processing is performed without directly calculating the expression (32).
[0222]
That is, for each control cycle, the following equation (from the standard atmospheric pressure Pa0, the standard intake pressure Pb0, and the atmospheric pressure Pa detected by the atmospheric pressure sensor 18 when the internal combustion engine 1 is started (starting mode)) 33), the parameter ratio / dpa (hereinafter referred to as the atmospheric pressure correction parameter ratio / dpa), that is, the value within the square root (√) of equation (32).
[0223]
[Expression 33]

[0224]
Then, the data table of FIG. 10 is prepared by calculating the square root for various values of the atmospheric pressure correction parameter ratio / dpa in advance, and the previously obtained atmospheric pressure correction parameter ratio / dpa is prepared. The square root is obtained as the atmospheric pressure correction coefficient kpa. Then, the opening degree command is corrected (multiplication correction) using the atmospheric pressure correction coefficient kpa.
[0225]
The atmospheric pressure correction coefficient kpa is “1” if the atmospheric pressure Pa detected by the atmospheric pressure sensor 18 is the standard atmospheric pressure Pa0, and the value decreases as the atmospheric pressure Pa increases.
[0226]
The contents described above are the basic contents of the process for compensating the intake air amount variation due to the atmospheric pressure in the atmospheric condition correction process.
[0227]
Next, the atmospheric temperature in the atmospheric conditions will be described. As is apparent from the equation (28), the intake air amount Gcyl of the combustion chamber 4 of the internal combustion engine 1 is affected by the coefficient Ci corresponding to the atmospheric density, and increases as the atmospheric density increases. Since the atmospheric density is lower as the atmospheric temperature is higher, the intake air amount Gcyl of the combustion chamber 4 is lower as the atmospheric temperature is higher.
[0228]
Therefore, in order not to change the intake air amount Gcyl of the combustion chamber 4 even if the atmospheric temperature changes, the opening degree command Θ should be corrected so that the higher the atmospheric temperature, the larger the bypass opening degree. Good.
[0229]
Therefore, in the present embodiment, the correction is made from the atmospheric temperature Ta detected by the atmospheric temperature sensor 17 when the internal combustion engine 1 is started (starting mode) based on a data table previously determined based on experiments as shown in FIG. A coefficient kta (hereinafter referred to as an atmospheric temperature correction coefficient kta) is obtained, and the opening degree command is corrected (multiplication correction) using the atmospheric temperature correction coefficient kta.
[0230]
In this case, in this embodiment, since the intake air amount at the standard atmospheric temperature Ta0 (for example, 25 ° C.) is used as a reference, the atmospheric temperature correction coefficient kta is “when the detected atmospheric temperature Ta is the standard atmospheric temperature Ta0”. 1 ”, the value increases as the atmospheric temperature Ta increases.
[0231]
This is the basic content of the process for compensating the intake air amount variation due to the atmospheric temperature in the atmospheric condition correction process.
[0232]
Next, the learning correction process will be described.
[0233]
In the present embodiment, as described above, in order to compensate for the variation in the intake air amount due to the structural factor, the SLD opening degree is controlled for each control cycle by the intake air amount F / B control correction process using the intake side adaptive SLD control. A correction amount i / sld is obtained, and the standard opening command value θ0 is corrected by the SLD opening correction amount i / sld. In this case, due to aging deterioration of the bypass valve 7 and the like, the actual intake air amount with respect to the standard opening command value θ0 varies in comparison with the original standard intake air amount corresponding to the standard opening command value θ0. In such a state that the accumulated intake air amount Qa (estimated accumulated intake air amount qair / pre) is not yet converged with respect to the target accumulated amount q (target accumulated intake air amount qair / cmd). The SLD opening correction amount i / sld becomes large, and the change with time also becomes large. For this reason, the opening command Θ determined by correcting the standard opening command value θ0 by the SLD opening correction amount i / sld, and thus the actual temporal change pattern of the intake air amount, is the initial stage of the FIRE mode. There is a case where a temporal change greatly deviating from the temporal change pattern of the standard opening command value θ0 (the target temporal change pattern of the intake air amount) may occur.
[0234]
However, particularly in the initial stage of the FIRE mode (a state immediately after the start of the internal combustion engine 1), the combustion state of the air-fuel mixture in the combustion chamber 4 tends to become unstable, and the pattern of temporal change of the opening degree command Θ is If a change greatly deviating from the temporal change pattern of the standard opening command value θ0 is exhibited, the combustion state and the emission state may be deteriorated.
[0235]
Therefore, in the learning correction processing in the present embodiment, learning for each control cycle is learned during operation in the FIRE mode, and correction for multiplying and correcting the standard opening command value θ0 over the entire period of the next FIRE mode operation. A coefficient kilearn (hereinafter referred to as a learning correction coefficient kilearn) is obtained. Then, by multiplying the standard opening command value θ0 by the learning correction coefficient kilearn, an opening command θ that exhibits a stable temporal change consistent with the temporal change pattern of the standard opening command value θ0 is generated.
[0236]
In this case, the learning correction coefficient kilearn is determined as follows in the present embodiment.
[0237]
That is, the actual intake air amount corresponding to the SLD opening correction amount i / sld is calculated from the SLD opening correction amount i / sld obtained for each control cycle during each FIRE mode operation by the following equation (34). Correction amount gair / sld (hereinafter referred to as SLD intake correction amount gair / sld) is obtained for each control cycle.
[0238]
[Expression 34]

[0239]
Here, the equation (34) is an equation similar to the equation (4) for obtaining the target intake air amount gair / cmd, that is, the actual intake air amount (per 1 TDC) corresponding to the standard opening command value θ0. Ga2 in formula (34) is the same as Ga2 in formula (4).
[0240]
Further, the SLD intake correction amount gair / sld is cumulatively added for each control cycle according to the following equation (35), so that an integrated value qair / sld of the SLD intake correction amount gair / sld (hereinafter, SLD integrated intake correction amount qair / sld).
[0241]
[Expression 35]

[0242]
Then, a value obtained by the following equation (36) from the ratio (qair / sld / qair / cmd) of the SLD integrated intake air correction amount qair / sld to the target integrated intake air amount qair / cmd is a basic value of the learning correction coefficient kilolen. vpskisld (hereinafter referred to as basic learning correction coefficient vpskisld) is obtained.
[0243]
[Expression 36]

[0244]
The basic learning correction coefficient vpskisld is basically calculated for each control cycle until the operation of the FIRE mode ends. However, when the interruption operation of the FIRE mode is performed, an ignition timing response correction process described later is performed. Since the calculation of the SLD opening correction amount i / sld is interrupted as described above, before the FIRE mode interruption operation or the ignition timing response correction process is started (the learning calculation end flag f The calculation of the basic learning correction coefficient vpskisld is completed until / flrnend is set to “1”.
[0245]
The smoothing calculation process (filtering process) according to the following equation (37) is performed on the basic learning correction coefficient vpskisld finally obtained during the operation of each FIRE mode in this way, so that the next FIRE is performed. The learning correction coefficient kilearn for correcting the standard opening command value θ0 during the mode operation is obtained.
[0246]
[Expression 37]

[0247]
Here, kilearn (j) in the equation (37) means a learning correction coefficient kilearn newly determined by the operation in the current FIRE mode, and kilearn (j-1) is determined by the operation in the previous FIRE mode. Means the corrected learning correction coefficient kilearn. Further, Cki in the equation (37) is a predetermined constant of “1” or less.
[0248]
In this case, in each operation of the FIRE mode, the FIRE elapsed time t / fire when the final basic learning correction coefficient vpskisld is obtained (this is STEP 5-5, 5-5 in the flowchart of FIG. 5). The basic learning correction coefficient vpskisld is not used to obtain (update) the learning correction coefficient kilearn when the parameter t / kil (set by 16 etc.) is less than the predetermined time. The learning correction coefficient kilearn is maintained. This is because the basic learning correction coefficient vpskisld obtained at a stage where the FIRE elapsed time t / fire is short is not reliable.
[0249]
The process described above is the basic content of the learning correction process.
[0250]
Next, the ignition timing response correction process will be described.
[0251]
In the present embodiment, as described above, when the warm-up of the internal combustion engine 1 proceeds to some extent (when the FIRE elapsed time t / fire increases to some extent), the rotational speed Ne of the internal combustion engine 1 due to the reduction of the friction of each part of the internal combustion engine 1. In order to prevent the ignition timing from being increased and to prevent the ignition timing from being excessively retarded by the ignition timing operation speed F / B control, the standard opening command value θ0 is gradually decreased. (See FIG. 7).
[0252]
However, the form of the reduction in friction accompanying the progress of warm-up of the internal combustion engine 1 is affected by various factors, and the reduction in the friction starts earlier than expected, or the degree of reduction in the friction is expected. It can be bigger than that.
[0253]
In such a case, even if the standard opening command value θ0 is gradually decreased as described above, it is impossible to sufficiently suppress the increasing tendency of the rotational speed Ne of the internal combustion engine 1. As a result, the ignition timing operated by the ignition timing operation rotational speed F / B control reaches the limit value on the retard side that can actually be operated. As a result, the rotational speed Ne is set to the target rotational speed ne / fire. It becomes impossible to perform feedback control.
[0254]
The ignition timing response correction process is a process performed to avoid such a situation, and the ignition timing command value determined as will be described later by the ignition timing operation rotational speed F / B control is a retarded limit value. When the value on the retarded angle side exceeds the predetermined threshold value slightly set on the advanced angle side until the state is resolved (the ignition timing command value returns to the advanced value above the threshold value). The standard opening command value θ0 is corrected to the decreasing side by a predetermined amount every control cycle.
[0255]
That is, for example, when the bypass opening is operated in accordance with the standard opening command value θ0 as shown in the upper part of FIG. 12, the ignition timing operation speed F / B control is performed as shown in the A12 region in the lower part of FIG. When the determined ignition timing command value iglog falls to the retarded angle side with respect to the threshold value IGX that is slightly larger than the retarded limit value IGLGG, the opening command Θ is compared with the standard opening command value θ0 as shown in FIG. As shown in the upper stage, θdec is decreased by a certain correction amount (hereinafter referred to as ignition timing responsive opening correction amount θdec) (referred to as Θ = θ0−θdec).
[0256]
In this case, the ignition timing responsive opening correction amount θdec is a predetermined amount for each control cycle until the ignition timing command value iglog returns to the advance side greater than or equal to the threshold value IGX as shown in the upper B12 region of FIG. It is increased by Δθdec (> 0, hereinafter referred to as an opening reduction unit amount Δθdec) (see the equation of STEP 15-9 in FIG. 15).
[0257]
Then, after the ignition timing command value iglog has returned to the advance side greater than the threshold value IGX as shown in the lower C12 region of FIG. 12, the ignition timing responsive opening correction is performed as shown in the upper D12 region of FIG. The amount θdec is held at the value at the time of return (the process of increasing the ignition timing responsive opening correction amount θdec by the opening decrease unit amount Δθdec is stopped), and the held ignition timing responsive opening correction amount θdec is The opening command Θ is corrected to be smaller than the standard opening command value θ0.
[0258]
In the present embodiment, the opening decrease unit amount Δθdec for increasing the ignition timing responsive opening correction amount θdec for each control cycle is determined in advance from the engine temperature Tw when the internal combustion engine 1 is started as shown in FIG. Determine based on the data table. In this case, in particular, when the engine temperature Tw at the start of the internal combustion engine 1 is in the high temperature region, the reduction in friction is more likely to occur than in the case where the engine temperature Tw is in the low / medium temperature region. In the table, the opening degree decrease unit amount Δθdec in the high temperature region of the engine temperature Tw is set to be larger than that in the low intermediate temperature region.
[0259]
The contents described above are the basic contents of the ignition timing response correction process.
[0260]
Based on the above description, the process of generating the bypass opening command value (opening command) θCMD performed by the intake air amount control stage 25 in STEP 4-5 in FIG. 4 will be described in detail.
[0261]
Referring to the flowchart of FIG. 14, in STEP 4-5, first, the value of FIRE execution enable / disable flag f / fireon set for each control cycle as described above in the condition determination process (see FIG. 15) of STEP 4-4. (Value set in the current control cycle) is determined (STEP 14-1).
[0262]
At this time, when f / fireon = 1, that is, when the operation mode is set to the FIRE mode, the N range basic used for obtaining the standard opening command value θ0 by the standard opening command value generation process. A value ifiret and a D range correction value iatfire, the atmospheric pressure correction coefficient kpa and the atmospheric temperature correction coefficient kta used for the atmospheric condition correction process, and the opening decrease unit amount Δθdec used for the ignition timing response correction process, respectively. (STEP 14-2) using a data table determined as described above.
[0263]
That is, the N range basic value ifiret and the D range correction value iatfire are obtained from the engine temperature Tw at the start of the internal combustion engine 1 acquired in the start mode processing (STEP 4-2) based on the data table of FIG.
[0264]
Further, from the atmospheric pressure Pa at the start of the internal combustion engine 1 acquired in the start mode process, and the predetermined standard atmospheric pressure Pa0 and standard intake pressure Pb0, the atmospheric pressure correction parameter ratio is obtained by the above equation (33). / dpa is calculated, and the atmospheric pressure correction coefficient kpa (= square root of the parameter ratio / dpa) is obtained from the parameter ratio / dpa based on the data table of FIG.
[0265]
Further, the atmospheric temperature correction coefficient kta is obtained from the atmospheric temperature Ta at the start of the internal combustion engine 1 acquired by the start mode process based on the data table of FIG.
[0266]
Further, from the engine temperature Tw at the start of the internal combustion engine 1 acquired in the start mode process, the opening decrease unit amount Δθdec is obtained based on the data table of FIG.
[0267]
Note that these processes in STEP 14-2 may be performed in advance in the start mode process.
[0268]
Next, the intake air amount operating means 25 calculates a preliminary opening degree command θi / fire obtained by subjecting the standard opening degree command value θ0 to the intake air amount F / B control correction process and the ignition timing response correction process. Processing is performed as follows (STEP 14-3).
[0269]
That is, referring to the flowchart of FIG. 15, the current shift position (current control cycle) of the automatic transmission detected by a sensor (not shown) is determined (STEP 15-1), and the shift position is in the N range. In this case, the basic value i / ftbl of the standard opening command value θ0 is set as the N range basic value ifiret obtained in STEP 14-2 (STEP 15-2). When the current shift position is the D range, a value obtained by adding the D range correction value iatfire obtained in STEP 14-2 to the N range basic value ifiret is set as a basic value i / ftbl (STEP 15-3). ).
[0270]
Next, the correction coefficient km / fire in the current control cycle is obtained from the current FIRE elapsed time t / fire based on the data table of FIG. 7 (STEP 15-4), and this correction coefficient km / fire is determined as STEP 15-2. Alternatively, the standard opening degree command value θ0 is obtained by multiplying the basic value i / ftbl determined in 15-3 (STEP 15-5). As a result, the standard opening command value θ0 for each control cycle in the FIRE mode is determined.
[0271]
Next, the intake air amount operating means 25 performs the process of calculating the intake air deviation Eq as shown in the flowchart of FIG. 16 (STEP 15-6).
[0272]
That is, referring to FIG. 16, first, an estimated intake air amount in the current control cycle is calculated from the current intake pressure Pb detected by the intake pressure sensor 16 and a predetermined value Ga1 in accordance with the equation (1). gair / pre (estimated value of intake air amount per 1 TDC) is obtained (STEP 16-1).
[0273]
Next, it is determined whether or not the fuel cut of the internal combustion engine 1 is in progress (STEP 16-2). If the fuel cut is not in progress, the current FIRE elapsed time t / fire further reaches the SLD correction limit time TISLDLMT. It is determined whether or not (STEP 16-3).
[0274]
At this time, if t / fire <TISLDLMT, the value of the estimated integrated intake air amount qair / pre (k) in the current control cycle is forcibly set to “0” (STEP 16-4). If t / fire ≧ TISLDLMT, the estimated cumulative intake air amount qair / pre (k) is obtained by accumulatively adding the estimated intake air amount gair / pre according to the equation (2) (STEP 16-5).
[0275]
When the fuel cut is being performed in STEP 16-2, the air sucked into the combustion chamber of the internal combustion engine 1 at that time does not contribute to the amount of heat given to the catalyst device 3 (the combustion of the internal combustion engine 1 during the fuel cut). Therefore, the estimated integrated intake air amount qair / pre (k) is maintained at the current value (STEP 16-6).
[0276]
In the FIRE mode (including the interruption operation) by the processing in STEPs 16-1 to 16-6 as described above, the fuel cut of the internal combustion engine 1 is being performed after the SLD correction limit time TISLDLMT has elapsed after the start. Except for the case, the estimated integrated intake air amount qair / pre (k) corresponding to the integrated value of the amount of heat actually applied to the catalyst device 3 is successively obtained.
[0277]
The estimated cumulative intake air amount qair / pre (k) is forcibly limited to “0” until the SLD correction limit time TISLDLMT elapses, that is, immediately after the internal combustion engine 1 is started. Become.
[0278]
In order to limit the value of the estimated integrated intake air amount qair / pre (k) until the SLD correction limit time TISLDLMT elapses in this way, for example, the estimated intake air amount until the SLD correction limit time TISLDLMT elapses. The value of gair / pre may be forcibly limited to “0”, and the calculation of equation (2) may be performed using the limited estimated intake air amount gair / pre. Even in this case, the value of the estimated integrated intake air amount qair / pre (k) is limited to “0” until the SLD correction limit time TISLDLMT elapses.
[0279]
After obtaining the estimated integrated intake air amount qair / pre (k) in this way, the intake air amount operating means 25 detects the standard opening command value θ 0 obtained in STEP 15-5 and the rotational speed sensor 14. The target intake air amount gair / cmd (target value of the intake air amount per 1 TDC) in the current control cycle is obtained from the current rotational speed Ne and a predetermined value Ga2 determined in advance by the above equation (4) ( (STEP 16-7).
[0280]
Further, the same determination as in STEP16-3 is performed (STEP16-8). At this time, if t / fire <TISLDLMT, the target integrated intake air amount qair / cmd (k) in the current control cycle is determined. Forcibly set to “0” (STEP 16-9). If t / fire ≧ TISLDLMT, the target integrated intake air amount qair / cmd (k) is obtained by cumulatively adding the target intake air amount gair / pre for each control cycle according to the equation (5) (STEP 16). -10).
[0281]
As a result of the processing in STEPs 16-7 to 16-10 as described above, the target integrated intake air amount qair / cmd (k) corresponding to the target value of the integrated value of the heat amount actually applied to the catalyst device 3 becomes the start of the FIRE mode. Thereafter, after the SLD correction limit time TISLDLMT elapses, the SLD correction limit time TISLDLMT is sequentially obtained, including during the interruption operation of the FIRE mode.
[0282]
Then, in the state immediately after the start of the internal combustion engine 1 until the SLD correction limit time TISLDLMT elapses, the value of the target integrated intake air amount qair / cmd (k) is the same as the estimated integrated intake air amount qair / pre (k). Forcibly limited to “0”.
[0283]
In order to limit the target integrated intake air amount qair / cmd (k) until the SLD correction limit time TISLDLMT elapses in this way, the target intake air amount gair is increased until the SLD correction limit time TISLDLMT elapses. The value of / cmd may be forcibly limited to “0”, and the calculation of the expression (5) may be performed using the limited target intake air amount gair / cmd.
[0284]
Thus, after obtaining the estimated integrated intake air amount qair / pre (k) and the target integrated intake air amount qair / cmd (k) in the current control cycle, the intake air amount operating means 25 Is calculated (qair / pre (k) −qair / cmd (k)) to obtain the intake air deviation Eq (k) in the current control cycle (STEP 16-11). Return.
[0285]
Returning to the explanation of FIG. 15, after obtaining the intake air deviation Eq as described above, the intake air amount operating means 25 next determines the value of the flag f / dec (STEP 15-7). This flag f / dec is a flag related to the ignition timing response correction process, and the ignition timing command value iglog is later than the threshold IGX (see FIG. 12) in the ignition timing command value iglog generation process described later. F / dec = 1 when the value is on the corner side, and f / dec = 0 when the command value iglog is a value on the advance side greater than the threshold value IGX (hereinafter, flag f / dec is called ignition timing determination flag f / dec). The ignition timing determination flag f / dec is initialized to “0” in the start mode process (STEP 4-2).
[0286]
At this time, if f / dec = 0 (the ignition timing command value iglog is on the advance side greater than the threshold value IGX), the intake air amount operating means 25 performs the intake air amount F / B control correction process. The process of calculating the SLD opening correction amount i / sld related to is performed as shown in the flowchart of FIG. 17 (STEP 15-8).
[0287]
That is, first, the current value of the FIRE interruption flag f / fpause is determined (STEP 17-1). At this time, when f / fpause = 1, that is, when the above-described interruption operation in the FIRE mode is performed, the intake air amount F / B control correction processing defines the attenuation rate of the intake air deviation Eq. After setting (initializing) the value of the pole pole / i to the predetermined lower limit value pole / i0 (see FIG. 9) (STEP 17-2), the routine immediately returns to the routine processing of FIG.
[0288]
Therefore, during the interruption operation in the FIRE mode, the SLD opening correction amount i / sld is held at the current value (the value before the start of the interruption operation).
[0289]
The SLD opening correction amount i / sld and the pole pole / i are initialized to “0” and “lower limit value pole / i0”, respectively, in the start mode process (STEP 4-2).
[0290]
On the other hand, if it is determined in STEP 17-1 that f / fpause = 0, that is, if the normal operation in the FIRE mode is performed, the intake air amount operating means 25 determines the current FIRE elapsed time t. From / fire, the pole table value pole / itbl in the current control cycle is obtained based on the data table of FIG. 9 (STEP 17-3).
[0291]
Next, basically, the pole table value pole / itbl obtained in STEP 17-3 is set as the value of the pole pole / i that defines the decay speed of the intake air deviation Eq. During the interruption operation of the FIRE mode, The pole pole / i set to the lower limit value pole / i0 ("-1" in the present embodiment) is gradually returned to the pole table value pole / itbl corresponding to the FIRE elapsed time t / fire after the interruption operation ends. Therefore, the following processing is performed.
[0292]
That is, a value (pole / i (k-1) + ΔPOLE / I) obtained by adding a predetermined unit increment value ΔPOLE / I (> 0) to the current pole / i value pole / i (k-1) It is compared with the pole table value pole / itbl obtained in STEP 17-3 (STEP 17-4). If pole / i (k-1) +. DELTA.POLE / I.gtoreq.pole / itbl, the pole table value pole / itbl obtained in STEP 17-3 is the value of the pole pole / i (k) in this control cycle. (STEP 17-5) If pole / i (k-1) + ΔPOLE / I <pole / itbl, the value of pole / i (k-1) + ΔPOLE / I is set to the pole in this control cycle. It is set as the value of pole / i (k) (STEP 17-6).
[0293]
After the interruption operation of the FIRE mode is completed by this processing, the pole increment of the pole pole / i from the lower limit value pole / i0 to the pole table value pole / itbl corresponding to the FIRE elapsed time t / fire is ΔPOLE / I. It will gradually return.
[0294]
If the FIRE mode is not interrupted immediately after the internal combustion engine 1 is started, the pole table pole / i is basically set to the pole table value pole / itbl, and the pole table value pole / itbl and In the same form, the value changes as the FIRE elapsed time t / fire increases (the unit increment value ΔPOLE / I is set so as to be like this).
[0295]
After setting the value of the pole pole / i in this way, the intake air amount operating means 25 further sets the value of the pole pole / i and the ignition timing operation speed F / B control which will be described in detail later. Pole pole / ig (This is a parameter that defines the decay rate of the deviation between the rotational speed Ne of the internal combustion engine 1 and the target rotational speed ne / fire by controlling the ignition timing operating rotational speed F / B). A value obtained by subtracting a predetermined minute amount ΔPOLE / IG (> 0) is compared (STEP 17-7).
[0296]
At this time, if pole / i <pole / ig−ΔPOLE / IG, the process proceeds to the next STEP 17-9, but if pole / i ≧ pole / ig−ΔPOLE / IG, the pole pole / i The value is forcibly reset to (pole / ig−ΔPOLE / IG) (STEP 17-8). That is, the value of the pole pole / i is always smaller than the pole pole / ig (<0) set as described later in the ignition timing operation speed F / B control (more precisely, 1> | pole / i |> | pole / ig |> 0). This is due to the following reason.
[0297]
That is, in this embodiment, the intake side adaptive SLD control (feedback control) is performed so that the estimated integrated intake air amount qair / pre converges to the target integrated intake air amount qair / cmd, and the rotational speed Ne of the internal combustion engine 1 is set to the target rotation. The ignition timing operation rotational speed F / B control performed so as to converge to the number ne / fire is performed independently of each other, while both controls are controls that affect the rotational speed Ne of the internal combustion engine 1. is there. In general, the response of the change in the intake air amount to the change in the bypass opening based on the intake side adaptive SLD control is the response of the change in the rotational speed Ne to the change in the ignition timing based on the ignition timing operation rotational speed F / B control. Slow compared to sex. Therefore, the decay rate of the intake deviation Eq related to the intake side adaptive SLD control is set to be higher than the decay rate of the deviation between the rotational speed Ne of the internal combustion engine 1 and the target rotational speed ne / fire by the ignition timing operation rotational speed F / B control. If the speed is increased, the two controls interfere with each other, and the rotational speed Ne of the internal combustion engine 1 may become unstable.
[0298]
For this reason, in the present embodiment, the pole pole / i related to the intake side adaptive SLD control is set so as to satisfy | pole / i |> | pole / ig | as described above. The damping speed of the intake air deviation Eq related to the control is made slower than the damping speed of the deviation between the rotational speed Ne of the internal combustion engine 1 and the target rotational speed ne / fire by the ignition timing operation rotational speed F / B control, so that both controls are performed. Avoid interfering with each other.
[0299]
Next, in STEP 17-9, the value of the pole pole / i is compared with “−1”, and when pole / i ≦ −1 (in such a case, it may be caused by the processing of STEP 17-8). After forcibly setting the value of pole pole / i to “−1” (STEP 17-10), the process proceeds to STEP 17-11.
[0300]
In STEP 17-11, the value of the pole pole / i determined as described above, the intake deviation Eq (k) in the current control cycle obtained as described above in STEP 15-6, and the intake air in the previous control cycle. From the deviation Eq (k-1), the above-described switching function σ1 is obtained by the above-described equation (10) (specifically, the equation is obtained by replacing the coefficient parameters s1 and s2 of equation (10) with “1” and “pole / i”, respectively). Find the value.
[0301]
Further, by using the value of the switching function σ1 and calculating the equations (18) and (19) (in this case, s1 = 1), the reaching law input Θrch and the adaptive law input in the intake side adaptive SLD control are performed. The value of Θadp is obtained (STEP 17-12), and the reaching law input Θrch and the adaptive law input Θadp are added to obtain the SLD opening correction amount i / sld in the current control cycle (STEP 17-13). Then, the process returns to the routine process of FIG.
[0302]
Returning to the description of FIG. 15, when f / dec = 1 in the determination of STEP 15-7, that is, the command value iglog of the current ignition timing is more retarded than the threshold IGX (see FIG. 12). If it is in the state, the intake air amount manipulating means 25 sets the ignition timing responsive opening correction amount θdec to the opening decrease unit amount Δθdec determined in STEP 14-2 in order to perform the ignition timing responsive correction processing. It is increased every control cycle (STEP 15-9).
[0303]
The ignition timing responsive opening correction amount θdec is initialized to “0” in the start mode process (STEP 3-2).
[0304]
Further, when the processing of STEP 15-9 is performed, the processing for obtaining the SLD opening correction amount i / sld is not performed, and the value of the SLD opening correction amount i / sld is determined by the ignition timing determination flag f / dec. The value before being set to “1” is held.
[0305]
Next, the intake air amount operating means 25 adds the current SLD opening correction amount i / sld to the standard opening command value θ0 obtained in STEP 15-5, and further, the current ignition timing responsive opening correction amount θdec. Is subtracted to calculate the preliminary opening command θi / fire (= θ0 + i / sld−θdec) based on the intake air amount F / B control correction process and the ignition timing response correction process (STEP 15-10).
[0306]
Next, the intake air amount operating means 25 determines the current value of the learning calculation end flag f / flrnend (STEP 15-11). This learning calculation end flag f / flrnend is set when the interruption operation is started (FIRE interruption flag f / fpause) while the FIRE mode is operating (when the FIRE execution enable / disable flag f / fireon is set to “1”). Is set to “1”) or when the ignition timing command value iglog is retarded from the threshold value IGX (see FIG. 12) and the ignition timing response correction process is started ( When the ignition timing determination flag f / dec is set to “1”), the basic learning correction coefficient vpskisld is set to “1” (step 5-17 in FIG. 5 and later). (See STEP 22-8 in FIG. 22).
[0307]
When f / flrnend = 0, that is, when the basic learning correction coefficient vpskisld is to be calculated, the basic learning correction coefficient vpskisld is calculated as described for the learning correction processing (STEP 15). -12).
[0308]
That is, for each control cycle, from the current SLD opening correction amount i / sld obtained in STEP 5-18, the current rotational speed Ne of the internal combustion engine 1, and the predetermined value Ga2, the equation (34) The SLD intake air correction amount gair / sld is obtained and cumulatively added by the equation (35) to obtain the SLD integrated air intake correction amount qair / sld. Then, by calculating the equation (36) from the SLD integrated intake air correction amount qair / sld and the target integrated intake air amount qair / cmd for each control cycle obtained in STEP 16-10 (see FIG. 16). The basic learning correction coefficient vpskisld is calculated.
[0309]
In this case, until the FIRE elapsed time t / fire reaches the SLD correction limit time TISLDLMT, qair / cmd = 0 as described above (at this time, the SLD integrated intake correction amount qair / sld is also “0”). Therefore, the value of the basic learning correction coefficient vpskisld is forcibly set to “1”.
[0310]
If f / flrnend = 1 in STEP15-11, that is, if the calculation of the basic learning correction coefficient vpskisld is to be terminated, the processing in STEP15-12 is omitted, and the basic learning correction coefficient vpskisld is omitted. Is not performed (in this case, the value of the basic learning correction coefficient vpskisld obtained in the control cycle before f / flrnend = 1 is determined as its final value).
[0311]
After performing the processing of STEPs 5-11 and 5-12 in this way, the intake air amount operating means 25 sets the value of the preliminary opening degree command θi / fire obtained in STEP 15-10 between a predetermined upper limit value and lower limit value. (When θi / fire> upper limit value or θi / fire <lower limit value, θi / fire is forcibly limited to an upper limit value and a lower limit value, respectively) after limit processing (STEP 15-13) ), The routine returns to the routine shown in FIG.
[0312]
Returning to the description of FIG. 14, after obtaining the preliminary opening degree command θi / fire as described above, the intake air amount operation means 25 sets the preliminary opening degree instruction θi / fire (= θ0 + i / sld−θdec). The atmospheric pressure correction coefficient kpa determined in STEP 14-2, the atmospheric temperature correction coefficient kta, and the learning correction coefficient kilearn determined at the end of the previous FIRE mode operation (the calculation of the learning correction coefficient kilearn will be described later) To determine the bypass opening command value θCMD in the current control cycle (STEP 14-4). Then, the process returns to the main routine process of FIG.
[0313]
The processing described above is processing for determining the bypass opening command value θCMD for each control cycle in the FIRE mode.
[0314]
On the other hand, if the value of the current FIRE execution enable / disable flag f / fireon is “0” as determined in STEP 14-1, that is, when the operation of the FIRE mode is not performed (this is basically Is the state after the FIRE elapsed time t / fire reaches the FIRE mode limit time TFIRELMT and the operation of the FIRE mode is terminated after the internal combustion engine 1 is started). The value in the previous control cycle of the execution flag f / fireon is determined (STEP 14-5).
[0315]
At this time, when the value of the FIRE execution enable / disable flag f / fireon in the previous control cycle is “1”, that is, when the operation of the FIRE mode has been performed until the previous control cycle (the state immediately after the end of the FIRE mode) ), The basic learning correction coefficient vpskisld finally obtained by the processing of STEP15-12 in the FIRE mode that has been performed until immediately before, as shown in the flowchart of FIG. A learning correction coefficient kilearn for correcting the command (see STEP 14-4) is determined (updated) (STEP 14-6).
[0316]
That is, referring to FIG. 18, first, the learning end time parameter t / kil representing the FIRE elapsed time t / fire when the calculation of the basic learning correction coefficient vpskisld described above is finally ended is set to a predetermined value. It is determined whether or not the value is TMKILLMT or more (STEP 18-1). Here, the learning end time parameter t / kil is the FIRE elapsed time t / fire when the learning calculation end flag f / flrnend is switched from “0” to “1” as shown in FIG. . That is, when the interruption operation in the FIRE mode is performed, it is the FIRE elapsed time t / fire until the start of the interruption operation. When the ignition timing response correction process is performed during the operation in the FIRE mode, the processing is performed. FIRE elapsed time t / fire until the start of. When the operation of the FIRE mode is completed without performing the interruption operation of the FIRE mode or the ignition timing response correction process, the FIRE elapsed time t / fire (this is usually the FIRE mode limit time TFIRELMT). Yes).
[0317]
In other words, the learning end time parameter t / kil is the FIRE process in which the calculation processing of the SLD opening correction amount i / sld and the correction of the standard opening command value θ0 corresponding thereto are continuously performed. Time is t / fire.
[0318]
If it is determined in STEP 18-1 that t / kil <TMKILLMT, the reliability of the finally obtained basic learning correction coefficient vpskisld is considered to be poor, so the learning correction coefficient kiilearn is maintained at the current value. Returning to the process of FIG.
[0319]
On the other hand, when t / kil ≧ TMKILLMT, that is, when the calculation processing of the SLD opening correction amount i / sld and the correction of the standard opening command value θ0 corresponding thereto are continuously performed for a certain long time. Is obtained from the basic learning correction coefficient vpskisld finally obtained in the operation of the FIRE mode performed until the previous control cycle by the smoothing calculation process (filtering process) of the above equation (37). j) is obtained (STEP 18-2). The learning correction coefficient kilearn (j) is limited to a value between a predetermined upper limit value and a lower limit value (when kilearn (j)> upper limit value or kilearn (j) <lower limit value, kilearn (j) Is forcibly limited to the upper limit value and the lower limit value) (STEP 18-3), then the routine returns to the routine process of FIG.
[0320]
Note that the initial value of the learning correction coefficient kilearn (j) is “1”. Further, the value of the learning correction coefficient kilearn (j) is stored and held in a nonvolatile memory such as an EEPROM so that it is not lost even when the operation of the system of this embodiment is stopped.
[0321]
Returning to the description of FIG. 14, when the value of the FIRE execution enable / disable flag f / fireon in the previous control cycle is “1” in STEP 14-5, the learning correction coefficient kilearn is updated as described above. If the value of the execution feasibility flag f / fireon in the previous control cycle is “0”, the learning correction coefficient kilearn has already been updated, so the processing of STEP 14-6 is omitted. That is, the update process of the learning correction coefficient kilearn is performed only in the control cycle immediately after the FIRE mode ends, and the updated learning correction coefficient kilearn is used to correct the opening degree command in the next operation of the FIRE mode. Is done.
[0322]
After performing the processing of STEPs 14-5 and 14-6 in this way, the intake air amount operation means 25 sets the opening degree command θCMD to the opening degree command for the normal mode (STEP 14-7). The opening command is basically smaller than the opening command θCMD in the FIRE mode, and is set to a required value for normal operation of the internal combustion engine 1.
[0323]
The process described above with reference to FIGS. 14 to 18 is the details of the process of generating the opening degree command (bypass opening degree command value) θCMD for each control cycle in STEP 4-5 of FIG. The opening degree command θCMD determined in this way is given to the bypass valve actuator 24. The bypass valve actuator 24 operates the opening degree of the bypass valve 7 in accordance with the given opening degree instruction θCMD.
[0324]
Next, the process of generating the ignition timing command value iglog in STEP 4-6 in FIG. 4 will be described.
[0325]
Here, before describing the specific contents of this process, the basic contents of this process will be described.
[0326]
In the system according to the present embodiment, the rotational speed Ne (actual rotational speed) of the internal combustion engine 1 tends to increase due to the increase control of the intake air amount as described above in the FIRE mode. Feedback control is performed to the required target rotational speed ne / fire by the number F / B control, and the ignition timing is corrected to the retard side (ignition timing operation rotational speed F / B control). In this case, as will be described later, it is preferable to appropriately variably set the rate of attenuation of the deviation between the rotational speed Ne and the target rotational speed ne / fire (this corresponds to a gain of feedback control).
[0327]
For this reason, in the ignition timing operation rotational speed F / B control in the system of the present embodiment, a response that can specify the rate of decay of the deviation as in the case of the intake air amount F / B control correction process described above. Sliding mode control (adaptive sliding mode control in this embodiment) which is designated type control is used.
[0328]
Processing of the ignition timing operation rotational speed F / B control using this adaptive sliding mode control (hereinafter referred to as ignition timing-side adaptive SLD control), that is, the ignition timing so as to converge the rotational speed Ne to the target rotational speed ne / fire. The processing algorithm for generating the command value iglog is constructed as follows.
[0329]
First, in the present embodiment, the control target of the ignition timing side adaptive SLD control is considered to be a system that generates data representing the rotational speed Ne (actual rotational speed) from data representing the ignition timing command value iglog, and this is discrete. Modeled in a system (discrete time system).
[0330]
In this case, in the present embodiment, the ignition timing command value iglog, the predetermined reference command value ig0, and the deviation DIG (= iglog−ig0, hereinafter referred to as the ignition timing deviation command value DIG) are data representing the ignition timing command value iglog. And a deviation DNE (= Ne−Ne0, hereinafter referred to as a deviation rotation speed DNE) between the rotation speed Ne and a predetermined reference rotation speed Ne0 is used as data representing the rotation speed Ne.
[0331]
In the present embodiment, the basic command value igbase (the command value of the ignition timing during normal operation other than the FIRE mode of the internal combustion engine 1) shown in FIG. 3 is used as the reference command value ig0 of the ignition timing ( ig0 = igbase). Therefore, the correction amount DIG shown in FIG. 3 is the ignition timing deviation command value DIG. In the present embodiment, the predetermined idling rotational speed NOBJ shown in FIG. 3 is used as the reference rotational speed Ne0 related to the rotational speed Ne (Ne0 = NOBJ).
[0332]
In this embodiment, using the ignition timing deviation command value DIG and the deviation rotation speed DNE, a model to be controlled by the ignition timing side adaptive SLD control is expressed as a discrete system model (38) In this embodiment, it is expressed by a second-order autoregressive model).
[0333]
[Formula 38]

[0334]
The model of the controlled object (hereinafter referred to as the rotational speed controlled object model) expressed by this equation (38) is the deviation rotational speed DNE (k + 1) in each control cycle as the output of the rotational speed controlled object model. It is expressed using the time series data DNE (K) and DNE (k-1) of the previous deviation speed DNE and the ignition timing deviation command value DIG (k) as the input of the speed control target model. is there.
[0335]
In Equation (38), the coefficients c1 and c2 related to the deviation speeds DNE (K) and DNE (k-1) and the coefficient d1 related to the ignition timing deviation command value DIG (k) These are model parameters that define the actual behavior characteristics of the target model. These model parameters c1, c2, and d1 include the behavior characteristics of the rotational speed control target model and the behavior characteristics of the actual control target expressed by the model. Identification is made in advance through experiments, simulations, etc. so as to match.
[0336]
In this case, since the rotational speed control target model is a discrete system model, various known identification algorithms (for example, the deviation rotational speed DNE (k + 1) generated on the rotational speed control target model, the actual deviation rotational speed, It is possible to identify the model parameters c1, c2, and d1 relatively easily using an algorithm that identifies the model parameters c1, c2, and d1 by the least square method so that the error between them is minimized.
[0337]
Based on the rotational speed control target model thus determined, the ignition timing side adaptive SLD control processing algorithm is constructed as follows.
[0338]
That is, in the ignition timing side adaptive SLD control, the switching function σ2 required for the sliding mode control is set to the deviation En = the deviation rotational speed DNE and its target value dne as in the case of the intake side adaptive sliding mode control described above. It is defined by a linear function of the following equation (39) using time series data En (k), En (k-1) for each control cycle of DNE-dne (hereinafter referred to as rotational speed deviation En) as variables. The target value dne of the deviation speed DNE (hereinafter referred to as the deviation target speed dne) is the deviation between the target speed ne / fire shown in FIG. 3 and the reference speed Ne0 (= idling speed NOBJ). (= Ne / fire-Ne0). Therefore, the rotational speed deviation En = DNE−dne is the same as the deviation (= Ne−ne / fire) between the rotational speed Ne (actual rotational speed) and the target rotational speed ne / fire.
[0339]
[39]

[0340]
Here, s3 and s4 in the equation (39) are coefficient parameters of each term of the switching function σ2, and are set so as to satisfy the following conditions.
[0341]
[Formula 40]

[0342]
In this embodiment, s3 = 1 is set for simplification. Further, the value of the coefficient parameter s4 (more generally, the value of s4 / s3 = pole / ig) is variably set, which will be described later.
[0343]
When the switching function σ2 is defined in this way, as in the case of the intake-side adaptive sliding mode control, the state quantity (a set of time series data En (k), En (k-1) of the rotational speed deviation En ( En (k), En (k-1)) converges on the switching line defined by the relational expression σ2 = 0, and when the convergence state is maintained, the state quantities (En (k), En (k-1) )) Can be converged very stably to an equilibrium point on the switching line σ2 = 0, that is, a point where En (k) = En (k-1) = 0, regardless of the influence of disturbance or the like.
[0344]
Further, in order to converge the deviation rotational speed DNE to the deviation target rotational speed dne (the rotational speed Ne converges to the target rotational speed ne / fire), an input to be given to the control target modeled by the equation (39) As in the case of the intake side adaptive sliding mode control, the control input generated by the ignition timing side adaptive SLD control, i.e., the ignition timing deviation command value DIG, is equivalent to the equivalent control input DIGeq, the reaching law input DIGrch and the adaptive law input DIGadp. It is the sum (see the following equation (41)).
[0345]
[Expression 41]

[0346]
These equivalent control input DIGeq, reaching law input DIGrch and adaptive law input DIGadp are given by the following equations (42) to (44), respectively, as in the case of the intake side adaptive sliding mode control.
[0347]
[Expression 42]

[0348]
[Equation 43]

[0349]
(44)

[0350]
In this case, the coefficient F3 (coefficient defining the reaching law gain) in the equation (43) is set in advance so as to satisfy the condition of the following equation (45), more preferably (45) ′.
[0351]
Further, the coefficient F4 (coefficient defining the gain of the adaptive law) in the equation (44) is set in advance so as to satisfy the condition of the following equation (46). In the equation (46), ΔT is a control cycle (control cycle).
[0352]
[Equation 45]

[0353]
[Equation 46]

[0354]
In the ignition timing side adaptive SLD control in the present embodiment, the equivalent control input DIGeq, the reaching law input DIGrch, and the adaptive law input DIGadp are respectively obtained from the equations (42) to (44) for each control cycle, and the sum of them is calculated. (Equation (41)) to obtain the ignition timing deviation command value DIG. Then, the ignition timing deviation command value DIG is added to the reference command value ig0, that is, the basic command value igbase during normal operation of the internal combustion engine 1, as shown in the following equation (47). Determine the value iglog.
[0355]
[Equation 47]

[0356]
In this case, the deviation target rotational speed dne necessary for obtaining the values of the equivalent control input DIGeq and the switching function σ2 is obtained as follows.
[0357]
That is, in the present embodiment, when the rotational speed Ne (actual rotational speed) of the internal combustion engine 1 reaches the set rotational speed (= NOBJ + NEFSLDS) shown in FIG. 3, or when the FIRE elapsed time t / fire is set in advance. The ignition timing operation speed F / B control is started when the predetermined value TSLDIGST is reached. The target rotational speed ne / fire of the internal combustion engine 1 in the FIRE mode is an elapsed time Δt / nfb (hereinafter referred to as rotational speed F / B elapsed time Δt / nfb) from the start of ignition timing operation rotational speed F / B control. ) According to the following equation (48).
[0358]
[Formula 48]

[0359]
Here, K / NE in the equation (48) is a predetermined value (> 0) that defines the temporal decrease degree (slope) of the target rotational speed ne / fire.
[0360]
In this case, when the calculation result on the right side of the equation (48) is lower than the idling speed NOBJ (when Δt / nfb> NEFSLDS / K / NE), the target speed ne / fire is set to the idling speed NOBJ. Secure to.
[0361]
As a result, the target rotational speed ne / fire gradually decreases linearly from the set rotational speed (= NOBJ + NEFSLDS) toward the idling rotational speed NOBJ after the ignition timing operation rotational speed F / B control is started. After reaching the rotational speed NOBJ, the idling rotational speed NOBJ is maintained.
[0362]
Therefore, in the present embodiment, the target rotational speed ne / fire (k) in each control cycle is obtained by the equation (48) from the rotational speed F / B elapsed time Δt / nfb in the control cycle, and this target rotational speed. By subtracting the idling rotation speed NOBJ from ne / fire (k), the deviation target rotation speed dne (k) (= ne / fire (k) −NOBJ) in each control cycle is obtained. The deviation target rotational speed dne (k) and its previous value dne (k-1) (= ne / fire (k-1) -NOBJ), that is, the deviation target rotational speed dne ( By using k-1), the value of the switching function σ2 (k) can be obtained by the above equation (39) for each control cycle. Then, by using the value of the switching function σ2, the reaching law input DIGrch and the adaptive law input DIGadp can be obtained from the above equations (43) and (44).
[0363]
In this embodiment, since the control cycle (TDC) is inversely proportional to the rotational speed Ne of the internal combustion engine 1, a time ΔT (∝1 / Ne) for one control cycle is obtained from the current rotational speed Ne, and the time By adding ΔT to the current rotation speed F / B elapsed time Δt / nfb, the rotation speed F / B elapsed time (= Δt / nfb + ΔT) in the next control cycle can be predicted (hereinafter, this prediction). The value is referred to as rotation speed F / B estimated elapsed time Δt / nfbpre). Then, this rotational speed F / B predicted elapsed time Δt / nfbpre is applied to the equation (48) (by substituting Δt / nfbpre for Δt / nfb on the right side of the equation (48)), the following equation (49) Thus, the target rotational speed ne / fire (k + 1) in the next control cycle can be obtained.
[0364]
[Formula 49]

[0365]
Further, by subtracting the idling rotational speed NOBJ from the target rotational speed ne / fire (k + 1), the deviation target rotational speed dne (k + 1) (= ne / fire (k + 1) in the next control cycle is obtained. -NOBJ) is required.
[0366]
Then, using the deviation target rotational speed dne (k + 1) obtained in this way and the deviation target rotational speeds dne (k) and dne (k-1) obtained as described above, By performing the calculation, the equivalent control input DIGeq can be obtained.
[0367]
In the present embodiment, during the interruption operation of the FIRE mode, the ignition timing operation rotational speed F / B control is interrupted, and the ignition timing command value iglog is changed to the basic command value igbase during normal operation of the internal combustion engine 1. return. In this case, the ignition timing deviation command value DIG value (absolute value) in the equation (47) is changed to a predetermined unit value dec / ig for each control cycle until the value becomes “0” after the interruption operation is started. By gradually decreasing it (by gradually approaching “0”), the ignition timing command value iglog is gradually returned to the basic command value igbase. When the interruption operation is released, the ignition timing operation rotational speed F / B control is immediately resumed.
[0368]
Further, in this embodiment, the operation of the FIRE mode is ended (the FIRE execution enable / disable flag f / fireon is changed from “1” to “0”), and the ignition timing is also changed when the system operation mode shifts to the normal mode. The command value iglog is gradually returned to the basic command value igbase.
[0369]
The contents described above are the basic contents of the ignition timing operation speed F / B control in the present embodiment.
[0370]
By the way, in the ignition timing side adaptive SLD control which is the response designation type control, the ratio (s4 / s3) of the coefficient parameters s3 and s4 of the switching function .sigma. The value of this ratio is referred to as pole pole / ig), and the damping speed of the rotational speed deviation En (= DNE−dne = Ne−ne / fire) can be designated. That is, as the absolute value of the pole pole / ig (= s4 / s3) is brought closer to "0" in a range smaller than "1", the decay rate of the rotational speed deviation En becomes faster. In this embodiment, as in the case of the intake side adaptive SLD control, s4 / s3 = pole / ig <0 is set in order to avoid vibrational attenuation of the rotational speed deviation En.
[0371]
In this embodiment, the damping speed of the rotational speed deviation En is variably set as follows using the response designation characteristic of the ignition timing side adaptive SLD control.
[0372]
That is, the change in the rotational speed Ne of the internal combustion engine 1 with respect to the change in the ignition timing tends to increase as the ignition timing is retarded. For this reason, in order to stably control the rotational speed Ne to the target rotational speed ne / fire, the damping speed of the rotational speed deviation En is delayed as the ignition timing being operated is retarded. It is considered preferable to increase the absolute value of pole pole / ig.
[0373]
Therefore, in this embodiment, for each control cycle, from the current ignition timing command value iglog (command value iglog determined in the previous control cycle), the pole pole is determined based on a predetermined data table as shown in FIG. The basic value pole / igtbl of / ig is obtained, and this basic value pole / igtbl is basically set as the value of the pole pole / ig. In this case, the basic value pole / igtbl (hereinafter referred to as ignition timing-corresponding pole basic value pole / igtbl) increases its absolute value | pole / igtbl | as the ignition timing command value iglog is retarded. (However, -1 <pole / igtbl <0).
[0374]
The ignition timing-corresponding pole basic value pole / igtbl is basically a value closer to “0” than the pole table value pole / itbl determined by the data table of FIG. 9 for the intake-side adaptive SLD control. Is set. In other words, the damping speed of the rotational speed deviation En defined by the ignition timing adaptive pole basic value pole / igtbl for the ignition timing adaptive SLD control is defined by the pole table value pole / itbl for the intake adaptive SLD control. The ignition timing pole basic value pole / igtbl is set so as to be faster than the decay rate of the intake air deviation Eq.
[0375]
Further, in the present embodiment, when the ignition timing operation rotational speed F / B control is started, the ignition timing is set so as to converge the rotational speed Ne to the target rotational speed ne / fire by the processing of the ignition timing side adaptive SLD control described above. If it is changed rapidly, the combustion state of the internal combustion engine 1 may be deteriorated. Therefore, in the initial stage of the ignition timing operation rotational speed F / B control, it is preferable to set the value of the pole pole / ig so as to delay the decay speed of the rotational speed deviation En.
[0376]
Therefore, in this embodiment, for each control cycle, from the rotation speed F / B elapsed time Δt / nfb (elapsed time since the ignition timing operation rotation speed F / B control was started), as shown in FIG. Based on the determined data table (time table), a correction coefficient kigt for correcting (multiplying correction) the ignition timing corresponding pole basic value pole / igtbl according to the rotational speed F / B elapsed time Δt / nfb is obtained, and this correction is performed. The ignition timing corresponding pole basic value pole / igtbl is corrected by multiplying a coefficient kigt (hereinafter referred to as a time corresponding correction coefficient kigt) by the ignition timing corresponding pole basic value pole / igtbl.
[0377]
In this case, in the data table of FIG. 21, the time-related correction coefficient kigt is the initial stage of the ignition timing operation rotational speed F / B control (until the rotational speed F / B elapsed time Δt / nfb reaches a predetermined value T / NFBX). Then, a value that corrects the ignition timing-corresponding pole basic value pole / igtbl in a direction in which the absolute value of the ignition timing-corresponding pole basic value pole / igtbl is slightly increased (a direction in which the decay rate of the rotational speed deviation En is decreased) ( > 1). In addition, at this time, the time-related correction coefficient kigt is set to a larger value as the rotational speed F / B elapsed time Δt / nfb is shorter, so that the ignition timing is increased as the rotational speed F / B elapsed time Δt / nfb is shorter. The ignition timing corresponding pole basic value pole / igtbl is corrected so that the absolute value of the corresponding pole basic value pole / igtbl becomes larger (the decay speed of the rotational speed deviation En becomes slower). Then, after the rotational speed F / B elapsed time Δt / nfb reaches the predetermined value T / NFBX, the time correspondence correction coefficient kigt is held at “1”. In this state, the ignition timing correspondence pole basic value pole / igtbl Is not corrected.
[0378]
Further, in the present embodiment, during the interruption operation of the FIRE mode, the ignition timing operation rotational speed F / B control is interrupted, and the feedback control of the rotational speed Ne is not performed. When restarting the number F / B control, the rotational speed of the internal combustion engine 1 may be significantly higher than the target rotational speed ne / fire. In such a state, since the rotational speed deviation En becomes large, the ignition timing command value iglog obtained as described above by the ignition timing operating rotational speed F / B control is suddenly excessively increased to the retard side. The combustion state of the internal combustion engine 1 may be deteriorated. Accordingly, when the ignition timing operation rotational speed F / B control is resumed by releasing the interruption operation in the FIRE mode, the rotational speed Ne of the internal combustion engine 1 is greatly separated from the target rotational speed ne / fire toward the high rotational side. In such a situation, the decay rate of the rotational speed deviation En is slowed, and consequently, the ignition timing deviation command value DIG that makes the ignition timing command value iglog excessively retarded is avoided. It is considered preferable.
[0379]
Therefore, in this embodiment, the ignition timing corresponding pole basic value pole / igtbl is determined from the current rotational speed Ne of the internal combustion engine 1 for each control cycle based on a predetermined data table as shown in FIG. A correction coefficient kigne (hereinafter referred to as a rotation speed corresponding correction coefficient kigne) for correction (multiplication correction) is obtained in accordance with. In this case, in the data table of FIG. 21, the correction coefficient kigne corresponding to the rotational speed basically has a higher rotational speed Ne (the rotational speed Ne is more separated from the target rotational speed ne / fire on the higher rotational side). A value that corrects the ignition timing corresponding pole basic value pole / igtbl in the direction of increasing the absolute value of the ignition timing corresponding pole basic value pole / igtbl (in the direction of decreasing the decay rate of the rotational speed deviation En) ( > 1). In this embodiment, the ignition timing corresponding pole basic value pole / igtbl based on the rotation speed corresponding correction coefficient kilon is corrected for a predetermined period XCNT after the restart of the ignition timing operation rotation speed F / B control (this is shown in the above figure). In order to perform only the countdown timer cnt / igvpl set in STEP 5-18 of No. 3), the correction coefficient kigne corresponding to the rotational speed is corrected by the following equation (50), and this corrected correction coefficient kignef (hereinafter referred to as rotation) The ignition timing corresponding pole basic value pole / igtbl is corrected by multiplying the ignition timing corresponding pole basic value pole / igtbl by a number correction correction coefficient kignef).
[0380]
[Equation 50]

[0381]
Here, the countdown timer cnt / igvpl in the equation (50) is always set to the predetermined period XCNT during the interruption operation (the state of the FIRE interruption flag f / fpause = 1) by STEP5-18 in FIG. . When the ignition timing operation rotational speed F / B control is resumed by releasing the interruption operation, the value of the countdown timer cnt / igvpl decreases from the value of the predetermined period XCNT by a predetermined value every control cycle, and finally After the value becomes “0”, the value is held. Therefore, the correction correction coefficient kignef corresponding to the rotational speed obtained by the equation (50) is a value (≧ 1) corresponding to the rotational speed Ne until a predetermined period XCNT elapses after the ignition timing operation rotational speed F / B control is restarted. However, after the predetermined period XCNT has elapsed, kignef = 1 (in this case, the ignition timing corresponding pole basic value pole / igtbl is not corrected by the rotation speed corresponding correction coefficient kignef).
[0382]
The countdown timer cnt / igvpl is initialized to “0” in the start-up mode process (STEP 3-2). During the FIRE mode interruption operation, the XCNT is set for a predetermined period after the interruption operation ends. The value is held at “0” except for the period until it elapses.
[0383]
In the present embodiment, for each control cycle, the ignition timing-corresponding pole basic value pole / igtbl obtained as described above is multiplied by the time-corresponding correction coefficient kigt and the rotational speed-corresponding correction correction coefficient kignef as shown in the following equation (51). This value is set as the final pole / ig value.
[0384]
[Formula 51]

[0385]
The time-related correction coefficient kigt is kitt> 1 only in the initial stage immediately after the start of the ignition timing operation speed F / B control, and kitt = 1 in other states. Further, the correction correction coefficient kignef corresponding to the rotational speed is kignef> 1 at the initial stage immediately after the cancellation of the interruption operation in the FIRE mode, and only when the rotational speed Ne is relatively high, and kignef = 1 in other cases. Therefore, the value of the pole pole / ig set by the equation (51) is normally the ignition timing corresponding pole basic value pole / igtbl.
[0386]
Based on the contents described above, the process for generating the ignition timing command value iglog performed by the ignition timing operation means 26 in STEP 4-6 in FIG.
[0387]
Referring to the flowchart of FIG. 22, in STEP 4-6, the ignition timing operating means 26 first determines the basic command value igbase of the ignition timing (STEP 22-1). In this case, the basic command value igbase is obtained from a current rotational speed Ne, the intake pressure Pb, the engine temperature Tw, the atmospheric temperature Ta, and the like of the internal combustion engine 1 using a predetermined map or arithmetic expression.
[0388]
Next, the ignition timing operating means 26 executes a process for determining the ignition timing deviation command value DIG as shown in the flowchart of FIG. 23 (STEP 22-2).
[0389]
That is, in the processing of STEP 22-2, first, the value of the current FIRE execution enable / disable flag f / fireon is determined (STEP 23-1).
[0390]
At this time, when f / fireon = 1, that is, in a state where the operation in the FIRE mode is to be performed, the ignition timing operation means 26 determines whether or not to perform the ignition timing operation rotational speed F / B control by the value “1”. ”And“ 0 ”are determined (step 23-2).
[0390]
The rotation speed F / B execution feasibility flag f / nefb is initialized to “0” in the start mode process (STEP 4-2).
[0392]
In STEP23-2, when f / nefb = 0, that is, when the ignition timing operation rotational speed F / B control is not yet performed, the value of the rotational speed F / B elapsed time Δt / nfb is set to “0”. (STEP23-3). The rotational speed F / B elapsed time Δt / nfb is f / nefb = 1 in STEP23-2, and time measurement is started from the control cycle in which the ignition timing operation rotational speed F / B control should be performed. If f / nefb = 1 in STEP23-2, the processing in STEP23-3 is omitted.
[0393]
Next, the ignition timing operation means 26 obtains the target rotational speed ne / fire (k) in the current control cycle and the target rotational speed ne / fire (k + 1) in the next control cycle (STEP 23-4). In this case, the target rotational speed ne / fire (k) in the current control cycle is obtained from the current rotational speed F / B elapsed time Δt / nfb by the above equation (48). Regarding the target rotational speed ne / fire (k + 1) in the next control cycle, the time ΔT of one control cycle ascertained from the current rotational speed Ne (detected value of the rotational speed sensor 14) of the internal combustion engine 1 is set. The rotational speed F / B predicted elapsed time Δt / nfbpre is obtained by adding to the current rotational speed F / B elapsed time Δt / nfb, and the rotational speed F / B predicted elapsed time Δt / nfbpre is obtained from the equation (49). The target rotational speed ne / fire (k + 1) is obtained.
[0394]
Note that the target rotational speed ne / fire (k) in this control cycle is the state where the rotational speed F / B elapsed time Δt / nfb is set to “0” in STEP 23-3 (f / nefb = 0 state). ) Is the set rotational speed (NOBJ + NEFSLDS).
[0395]
Next, the ignition timing operation means 26 determines whether or not the current rotational speed Ne of the internal combustion engine 1 is equal to or higher than the current target rotational speed ne / fire (k) (STEP 23-5), and the current FIRE elapsed time t / fire. It is sequentially determined whether or not the value becomes equal to or greater than a predetermined value TSLDIGST (a constant value) (STEP 23-6).
[0396]
When any of the conditions of STEP23-5 and 23-6 is satisfied, the value of the rotation speed F / B execution feasibility flag f / nefb is set to “1” (STEP23-7). As a result, when the rotation speed Ne of the internal combustion engine 1 reaches the set rotation speed (NOBJ + NEFSLDS) or when the FIRE elapsed time t / fire reaches the predetermined value TSLDIGST after the operation in the FIRE mode is started. The number F / B execution feasibility flag f / nefb is set to “1”, and the ignition timing operation rotational speed F / B control can be performed.
[0397]
If neither of the conditions of STEP23-5 and 23-6 is satisfied, the processing of STEP23-7 is omitted and the value of the rotation speed F / B execution enable / disable flag f / nefb is maintained at “0”. . Further, the rotation speed F / B execution feasibility flag f / nefb is not returned to “0” during the operation of the FIRE mode after being set to “1”.
[0398]
Next, the ignition timing operation means 26 calculates the rotational speed deviation En (in the current control cycle from the current rotational speed Ne of the internal combustion engine 1 and the current target rotational speed ne / fire (k) obtained in STEP 23-4. k) (= Ne-ne / fire (k)) is calculated (STEP 23-8).
[0399]
Next, the ignition timing operation means 26 determines the value of the FIRE interruption flag f / fpause (STEP 23-9). If f / fpause = 0, that is, if the FIRE mode is not interrupted, the ignition timing operation means 26 further rotates. The value of the number F / B execution feasibility flag f / nefb is determined (STEP 23-10).
[0400]
At this time, if f / nefb = 0, the ignition timing deviation command value DIG (k) in the current control cycle is set to “0” (STEP 23-11), and the routine returns to the routine processing of FIG.
[0401]
On the other hand, if f / nefb = 1 in STEP23-10, the ignition timing deviation command value DIG (k) for controlling the ignition timing operation speed F / B is calculated (STEP23-12).
[0402]
The ignition timing deviation command value DIG (k) is calculated as follows.
[0403]
That is, referring to the flowchart of FIG. 24, first, the ignition timing correspondence is determined from the current command value iglog (k-1) (command value determined in the previous control cycle) of the ignition timing according to the data table of FIG. The basic pole value pole / igtbl is obtained (STEP 24-1).
[0404]
Further, the time corresponding correction coefficient kitt is obtained from the current rotational speed F / B elapsed time Δt / nfb by using the data table of FIG.
[0405]
Further, the rotational speed correspondence correction coefficient kilone is obtained from the current rotational speed Ne of the internal combustion engine 1 by the data table of FIG. 21 (STEP 24-3), and further, the rotational speed correspondence correction coefficient kilone and the countdown timer cnt / The correction correction coefficient kignef corresponding to the rotational speed is obtained from the current value of igvpl and a predetermined period XCNT predetermined as an initial value of the countdown timer cnt / igvpl after completion of the interruption operation in the FIRE mode by the above equation (50) ( (STEP 24-4).
[0406]
Then, the ignition timing corresponding pole basic value pole / igtbl obtained in STEP 24-1 is multiplied by the time correspondence correction coefficient kigt and the rotational speed correspondence correction correction coefficient kignef obtained in STEP 24-2 and 24-4, respectively (formula (51) The value of pole pole / ig in the current control cycle is determined (STEP 24-5).
[0407]
Next, the ignition timing manipulating means 26 determines the rotational speed deviations En (k) and En (k-1) obtained in STEP 23-8 in the current and previous control cycles, and the pole pole / From the current value of ig, the switching function σ2 (k) in the current control cycle is obtained by the calculation of the equation (39) (STEP 24-6). In this case, the coefficient parameters s3 and s4 in the equation (39) are "1" and "pole / ig", respectively.
[0408]
Next, the ignition timing manipulating means 26 obtains the equivalent control input DIGeq, the reaching law input DIGrch, and the adaptive law input DIDadp in the current control cycle from the equations (42) to (44), respectively (STEP 24-7).
[0409]
More specifically, with respect to the equivalent control input DIGeq, first, the target rotational speeds ne / fire (k) and ne / fire (k + 1) obtained in STEP 23-4 in the current control cycle and the previous control are determined. From the target rotational speed ne / fire (k-1) obtained in STEP23-4 in the cycle, those target rotational speeds ne / fire (k), ne / fire (k + 1), ne / fire (k-1) ) With respect to the reference rotational speed Ne0 (= idling rotational speed NOBJ), that is, the deviation rotational speeds dne (k), dne (k + 1), and dne (k-1). The deviation rotational speeds dne (k), dne (k + 1), dne (k-1) and the rotational speed deviations En (k), En obtained in STEP23-8 in the current and previous control cycles, respectively. The equivalent control input DIGeq (k) is obtained by performing the calculation of the equation (42) using (k-1) and the current value of the pole pole / ig obtained in STEP 24-5. In this case, the coefficient parameters s3 and s4 in the equation (42) are "1" and "pole / ig", respectively. The model parameters c1, c2, and d1 in the equation (42) are predetermined values identified in advance for the rotation speed control target model (see the equation (38)).
[0410]
Regarding the reaching law input DIGrch, the reaching law input DIGrch (k) is obtained by performing the calculation of the equation (43) using the value of the switching function σ2 (k) obtained in STEP24-6. In this case, the coefficient parameter s3 in the expression (43) is “1”, and the coefficient F3 in the expression (43) is a predetermined value set in advance so as to satisfy the condition of the expression (45) or (45) ′. Value.
[0411]
For the adaptive law input DIDadp, the integrated value Σσ2 of the switching function σ2 is obtained by cumulatively adding the value of the switching function σ2 obtained for each control cycle in STEP 24-6 for each control cycle. Then, the adaptive law input DIDadp is obtained by performing the calculation of the equation (44) using the integrated value Σσ2. In this case, the coefficient parameter s3 in the equation (44) is "1", and the coefficient F4 in the equation (44) is a predetermined value set in advance so as to satisfy the condition of the equation (46).
[0412]
After obtaining the equivalent control input DIGeq, the reaching law input DIGrch, and the adaptive law input DIDadp as described above, the ignition timing manipulating means 26 calculates their sum in accordance with the equation (41), so that the current control cycle The ignition timing deviation command value DIG (k) is obtained at (STEP 24-8), and the routine returns to the routine processing of FIG.
[0413]
Returning to the explanation of FIG. 23, after calculating the ignition timing deviation command value DIG (k) in STEP23-12 as described above, the ignition timing operating means 26 determines the value of the ignition timing deviation command value DIG (k). Is limited to a value between the predetermined upper limit value and lower limit value (DIG (k)> upper limit value, or DIG (k) <lower limit value, the value of DIG (k) is forcibly set to the upper limit value. After the limit process is performed (STEP 23-13), the routine returns to the routine process of FIG.
[0414]
When f / fpause = 1 in STEP23-9, that is, when the FIRE mode is to be interrupted, the ignition timing operation rotational speed F / B control is interrupted and the ignition timing command value is set. In order to gradually return the iglog to the basic command value igbase (the ignition timing deviation command value DIG is gradually returned to “0”), the unit value dec / ig (> 0. Hereinafter, the ignition timing return unit value dec / ig is determined (STEP 23-14).
[0415]
Further, when f / fireon = 1 in STEP 23-1, that is, when the operation of the FIRE mode is terminated or not performed, the ignition timing command value iglog is gradually changed to the basic command value igbase. In order to return, the ignition timing return unit value dec / ig is determined (STEP 23-15).
[0416]
In this case, in this embodiment, in STEP23-15, the ignition timing return unit value dec / ig is set to a predetermined value (a constant value), but the ignition timing set in STEP23-14 during the interruption operation of the FIRE mode. The return unit value dec / ig is determined to be a value proportional to the current opening degree of the throttle valve 5 (the ignition timing return unit value dec / ig increases as the opening degree of the throttle valve 5 increases).
[0417]
This is due to the following reason. That is, the interruption operation in the FIRE mode is basically performed in a situation where the vehicle travels or the internal combustion engine 1 is overrun by operating the accelerator pedal of the vehicle. At this time, the opening degree of the throttle valve 5 is operated to the opening degree corresponding to the accelerator operation amount Ap as described above, and the required operating performance of the internal combustion engine 1 is ensured in a situation where the opening degree of the throttle valve 5 is large. Therefore, it is considered preferable to return the normal ignition timing (which corresponds to the basic command value igbase) as soon as possible. For this purpose, in STEP23-14, the ignition timing return unit value dec / ig is determined to be a value proportional to the opening of the throttle valve 5.
[0418]
In this case, the opening degree of the throttle valve 5 required for determining the ignition timing return unit value dec / ig is an opening degree command value given from the controller 2 to the throttle valve actuator 23 or an opening degree by a sensor (not shown). Use the detected value.
[0419]
After determining the ignition timing return unit value dec / ig in this way, the ignition timing operating means 26 then selects the current ignition timing deviation command value DIG (this is the ignition timing determined in the previous control cycle). It is determined whether or not the timing deviation command value DIG (k-1) is smaller than "0", that is, whether or not it is a retarded value (STEP 24-16).
[0420]
When DIG (k-1) <0 (when DIG (k-1) is a retarded value), the current ignition timing deviation command value DIG (k-1) is added to the STEP24- A value (DIG (k-1) + dec / ig) obtained by adding the ignition timing return unit value dec / ig determined in 14 or 24-15 is determined as the ignition timing deviation command value DIG (k) in the current control cycle ( (STEP 24-17), the routine returns to the routine shown in FIG. In this case, if the upper limit value of the ignition timing deviation command value DIG (k) is “0” and (DIG (k−1) + dec / ig) is greater than “0”, the ignition timing The value of the deviation command value DIG (k) is forcibly set to “0”.
[0421]
On the other hand, when DIG (k-1) ≧ 0 in STEP24-16 (effectively, when DIG (k-1) = 0), the value of the FIRE executable flag f / fireon is determined (STEP24- 18) When f / fireon = 1 (in this case, the FIRE mode is being interrupted), the ignition timing deviation command value DIG (k) in the current control cycle is set to “0” (STEP 24-19) ), The routine returns to the routine shown in FIG.
[0422]
During the interruption operation in the FIRE mode (f / fireon = 1 and f / fpause state), the integrated value Σσ2 of the switching function σ2 is held at the value immediately before the start of the interruption operation.
[0423]
Further, when f / fireon = 0 in STEP24-18, that is, when the FIRE mode is finished, the values of the FIRE interruption flag f / fpause and the rotation speed F / B execution feasibility flag f / nefb are set to “0”. To the initial value of ignition timing deviation command value DIG (k), switching function σ2 value and its integrated value Σσ2, equivalent control input DIGeq, reaching law input DIGrch, adaptive law input DIGadp, etc. (STEP 24-20), the routine returns to the routine shown in FIG.
[0424]
Returning to the description of FIG. 22, as described above, after executing the process of determining the ignition timing deviation command value DIG in STEP 22-2, the ignition timing operating means 26 sets the STEP 22 to the basic command value igbase determined in STEP 22-1. The ignition timing command value iglog in the current control cycle is obtained by adding the ignition timing deviation command value DIG determined in -2 (STEP 22-3).
[0425]
Next, the ignition timing manipulating means 26 has a limit value IGLGG on the retarded side of the ignition timing shown in FIG. 12 with respect to the ignition timing response correction processing of the intake air amount manipulating means 25 described above, and is slightly advanced from the limit value IGLGG. Is determined (STEP 22-4).
[0426]
In this case, the retard side limit value IGLGG of the ignition timing is obtained from the engine temperature Tw of the internal combustion engine 1 based on a data table (not shown), and at the ignition timing on the advance side greater than the limit value IGLGG, the internal combustion engine 1 It is determined that normal operation can be performed. The threshold value IGX is set to a value obtained by adding a predetermined value (a constant value) to the limit value IGLGG (a value on the advance side by a predetermined value from the limit value IGLGG). The limit value IGLGG is a constant value when the engine temperature Tw of the internal combustion engine 1 is in a normal temperature range, but is normal when the engine temperature Tw is in a fairly low temperature range. It is set to a value on the advance side than in the temperature range.
[0427]
After the ignition timing retard limit value IGLGG and the threshold value IGX are thus determined, the ignition timing operating means 26 compares the command value iglog obtained in STEP 22-3 with the threshold value IGX (STEP 22-5). ). At this time, when the command value iglog is equal to the threshold value IGX or a value on the advance side of the threshold value IGX (iglog ≧ IGX), the ignition timing response correction process is not performed, so the ignition timing determination is performed. The value of the flag f / dec (see STEP 15-7 in FIG. 15) is set to “0” (STEP 22-6), and the process proceeds to STEP 22-9.
[0428]
If iglog <IGX in STEP 22-5, the value of the ignition timing determination flag f / dec is set to "1" to perform the ignition timing response correction process (STEP 22-7), and further the intake The value of the learning calculation end flag f / flrnend related to the learning correction process by the air amount operation means 25 is determined (STEP 22-8). The ignition timing operating means 26 holds the current value of the FIRE elapsed time t / fire as the value of the learning end time parameter t / kil only when f / flrnend = 0 (STEP 22-9 Further, the value of the learning calculation end flag f / flrnend is set to “1” in order to end the calculation processing of the basic learning correction coefficient vpskisld (STEP 22-10).
[0429]
Next, the ignition timing manipulating means 26 compares the command value iglog obtained in STEP 22-3 with the retard side limit value IGLGG determined in STEP 22-4 (STEP 22-11), and if iglog ≧ IGLGG, that is, the command When the value iglog is within the retarded limit value IGLGG, the process returns to the main routine of FIG.
[0430]
On the other hand, when iglog <IGLGG, that is, when the command value iglog obtained in STEP 22-3 exceeds the retard side limit value IGLGG and becomes the retard side value, the command value iglog is forced. Therefore, it is set to the limit value IGLGG (STEP 22-12). At this time, the integrated value Σσ2 of the value of the switching function σ2 is forcibly held (held) at the value obtained in the control cycle immediately before iglog <IGLGG (STEP 22-13), and then the main routine of FIG. Return to processing. Thus, the value of the integrated value Σσ2 is held in a state where the ignition timing command value iglog is forcibly set to the limit value IGLGG, and the calculation of the ignition timing deviation command value DIG is continued by the processing of STEP23-12. This is because the integrated value Σσ2 of the value of the switching function σ2 and thus the value (absolute value) of the adaptive law input DIGadp becomes excessively large.
[0431]
The processing described with reference to FIGS. 19 to 24 is the details of the processing for generating the command value iglog of the ignition timing for each control cycle in STEP 4-6 of FIG. The command value iglog is given to the ignition device 21. The ignition device 21 operates the ignition timing of the internal combustion engine 1 according to the given command value iglog.
[0432]
By the operation of the system of the present embodiment described above, in the operation of the FIRE mode, the intake air amount of the combustion chamber 4 of the internal combustion engine 1 is increased by the operation of the bypass opening as compared with the normal idling operation. In parallel with this, the rotational speed Ne of the internal combustion engine 1 is converged to the required target rotational speed ne / fire (finally the idling rotational speed NOBJ) by the ignition timing operation rotational speed F / B control. The ignition timing is operated to the retard side. As a result, the amount of heat given to the catalyst device 3 by the exhaust gas generated by the combustion of the air-fuel mixture by the internal combustion engine 1 is greater than that during normal idling operation, and the catalyst device 3 is heated and activated early. At the same time, the rotational speed Ne of the internal combustion engine 1 that tends to increase as the intake air amount increases can be maintained at an appropriate rotational speed.
[0433]
In this case, regarding the increase of the intake air amount, the estimated integrated intake air amount qair / pre corresponding to the integrated value of the instantaneous instantaneous heat amount actually given to the catalyst device 3 after the start of the increase is used as the catalyst device 3. Command of the bypass opening by the intake air amount F / B control correction process using the intake side adaptive SLD control so as to converge to the target integrated intake air amount qair / cmd corresponding to the target integrated value of the heat amount to be actually applied to The value is corrected.
[0434]
For example, as shown in the middle part of FIG. 25, the target integrated intake air amount qair / cmd in which the estimated integrated intake air amount qair / pre is determined according to the standard opening command value θ0 due to variations in the intake air amount due to the structural factors. When an error (intake deviation Eq) occurs, the opening degree command Θ (the command value of the bypass opening degree) is, as shown in the upper part of FIG. Only i / sld is corrected. As a result of this correction, the estimated integrated intake air amount qair / pre is converged to the target integrated intake air amount qair / cmd as shown in the middle of FIG. The target value can be followed. Thereby, the variation in the intake air amount due to the structural factor can be compensated, and consequently, the variation in the temperature raising mode of the catalyst device 3 due to the structural factor can be eliminated.
[0435]
Further, in the present embodiment, the opening degree command Θ is multiplied and corrected by the atmospheric pressure correction coefficient kpa and the atmospheric temperature correction coefficient kta obtained by the atmospheric condition correction process. That is, as the atmospheric pressure Pa is lower, the opening degree command Θ is corrected to the increasing side, and as the atmospheric temperature is higher, the opening degree command Θ is corrected to the increasing side. As a result, variations in the intake air amount due to atmospheric conditions can be compensated, and accordingly, variations in the temperature raising mode of the catalyst device 3 due to the atmospheric conditions can also be eliminated.
[0436]
As a result, the temperature raising mode of the catalyst device 3 in the FIRE mode can be matched with a desired temperature raising mode (in this embodiment, the engine temperature Tw at the start of the internal combustion engine 1 and the automatic transmission during the FIRE mode operation). If the shift position of the machine is constant, the temperature increase mode of the catalyst device 3 by the operation of the FIRE mode is almost the same except when the operation of interruption of the FIRE mode is performed), according to the operation of the FIRE mode The desired temperature increase and activation of the catalyst device 3 can be reliably performed.
[0437]
Further, the standard opening command value θ0, and thus the target intake air amount gair / cmd and the target integrated intake air amount qair / cmd, start of the internal combustion engine 1 corresponding to the temperature state of the catalyst device 3 when the internal combustion engine 1 is started. The internal combustion engine is set so as to correspond to the engine temperature Tw at the time (basically, the higher the engine temperature Tw, the smaller the target intake air amount gair / cmd and the target integrated intake air amount qair / cmd). Thus, the temperature rise and activation of the catalyst device 3 suitable for the temperature state of the catalyst device 3 at the time of starting 1 can be performed. That is, the temperature rise mode (degree of increase in temperature with time) of the catalyst device 3 in the FIRE mode depends on the temperature state of the catalyst device 3 at the start of the internal combustion engine 1, but is the final by the operation of the FIRE mode. The temperature state of the typical catalyst device 3 can be brought to an appropriate temperature state upon activation of the catalyst device 3.
[0438]
Further, the target integrated intake air amount qair / cmd defined by the standard opening command value θ0 corresponds to the intake air amount that should be sucked into the combustion chamber 4 of the internal combustion engine 1 for each control cycle, that is, the target intake air. The amount gair / cmd is set so that it gradually decreases after the FIRE elapsed time t / fire reaches a predetermined value t2 (see FIG. 7) (after the internal combustion engine 1 has warmed up to some extent). Is done. For this reason, it is possible to prevent the rotational speed Ne of the internal combustion engine 1 from increasing as the warm-up of the internal combustion engine 1 decreases even if the friction of each part thereof decreases. As a result, it is possible to prevent the ignition timing of the internal combustion engine 1 from being excessively retarded by the ignition timing operation F / B control.
[0439]
Further, in the present embodiment, when the ignition timing command value iglog determined by the ignition timing operation rotational speed F / B control exceeds a threshold value IGX that is close to the retard limit IGLGG that can operate the ignition timing. The intake air amount F / B control correction process is interrupted, and the opening degree command Θ is reduced in a feed-forward manner by the ignition timing response correction process. For this reason, even when the decrease in the friction of the internal combustion engine 1 is larger than expected or occurs early, the increase in the rotational speed Ne of the internal combustion engine 1 is suppressed, and the ignition timing is limited to the retard side. As the value IGLGG is reached, it is possible to avoid a situation in which the operation is excessively retarded.
[0440]
In the present embodiment, the intake air amount F / B control correction process uses sliding mode control that is not easily affected by disturbance, modeling error of the control target, or disturbance, and in particular, to eliminate the influence of disturbance or the like as much as possible. By using adaptive sliding mode control (intake-side adaptive SLD control) using the above-mentioned adaptive law, the estimated integrated intake air amount qair / pre is converged to the target integrated intake air amount qair / cmd, and is given to the catalyst device 3 as a result. Control for converging the integrated value of the heat amount to the target value can be performed with high stability. As a result, the desired temperature increase and activation of the catalyst device 3 can be performed more reliably.
[0441]
In the intake side adaptive SLD control of the intake air amount F / B control correction process, the control target is a system that generates the integrated intake air amount Qa from the opening degree command Θ. Therefore, the ignition timing side adaptive SLD control algorithm can be constructed simply and suitable for computer processing as compared with the case where the control object is expressed by a continuous system model.
[0442]
In particular, by modeling the control target of the intake side adaptive SLD control in a discrete system, the switching function σ1 used for the intake side adaptive SLD control is used only for the intake deviation Eq without using the change rate of the intake deviation Eq. The time series data can be configured. As a result, it is possible to increase the reliability of the value of the switching function σ1 necessary for obtaining the SLD opening correction amount i / sld, and hence the reliability of the intake side adaptive SLD control.
[0443]
Further, in the intake side adaptive SLD control, the feedback term Θeq / fb of the equivalent control input Θeq is omitted, and the standard opening command value θ0 corresponding to the feedforward term Θeq / ff is set to the equivalent control input Θeq in this embodiment. Thus, the SLD opening correction amount i / sld can be obtained by a simple algorithm without using the model parameters a1, a2, and b1 of the model to be controlled (intake control target model).
[0444]
Of course, the SLD opening correction amount i / sld may be obtained using the model parameters a1, a2, and b1, but even in this case, the control target of the intake-side adaptive SLD control is also determined. Since the model is constructed in a discrete system, it is possible to accurately identify the values of the model parameters a1, a2, and b1 by using a known identification algorithm.
[0445]
Further, in the intake air amount F / B control correction process, as shown in the middle part of FIG. 25, the estimated integrated intake air amount qair until the SLD correction limit time TISLDLMT immediately after the start of the internal combustion engine 1 elapses. The value of / pre and the target integrated intake air amount qair / cmd are forcibly held at “0”, and the value of the SLD opening correction amount i / sld is also held at “0”.
[0446]
For this reason, immediately after the internal combustion engine 1 is started, the intake air amount F / B control correction process is not performed until the SLD correction limit time TISLDLMT elapses, and the bypass opening is set to the standard opening command value θ0 (more accurately. Is operated in a feed-forward manner mainly with a standard opening command value θ0 multiplied by an atmospheric pressure correction coefficient kpa, an atmospheric temperature correction coefficient kta, and a learning correction coefficient kilearn (see the upper part of FIG. 25). reference).
[0447]
Therefore, immediately after the internal combustion engine 1 is started, the intake air amount is smoothly increased in the same manner as the standard opening command value θ0 is increased. As a result, the combustion of the internal combustion engine 1 immediately after the start is started. The state can be smoothly stabilized and a good emission state of the internal combustion engine 1 can be ensured.
[0448]
Further, when the intake air amount F / B control correction process is actually started, as shown in the lower part of FIG. 25, the value of the pole pole / i related to the intake-side adaptive SLD control is set to a predetermined FIRE elapsed time t / fire. By gradually increasing from the “−1” side (| pole / i | is gradually decreased) until the value TPOLEX is reached, the decay rate of the intake deviation Eq is reduced to the predetermined value TPOLEX. Until reaching the predetermined value TPOLEX, it is made slower than that.
[0449]
As a result, the start of the intake air amount F / B control correction process avoids a situation in which the opening degree command Θ, and thus the intake air amount, changes abruptly. This also prevents the internal combustion engine in the initial stage after starting. Thus, it is possible to ensure a good emission state of the internal combustion engine 1 while maintaining the stability of the combustion state of 1.
[0450]
In the present embodiment, the learning correction processing learns the SLD opening correction amount i / sld by the intake air amount F / B control correction processing for each operation in the FIRE mode, and the SLD opening correction amount i / sld is calculated. A learning correction coefficient kilearn based on the ratio of the SLD integrated intake air correction amount qair / sld integrated during the operation of the FIRE mode to the target integrated intake air amount qair / cmd (basically, the basic learning correction coefficient vpskisld) Ask for. Then, in the next operation of the FIRE mode, the learning instruction correction coefficient kilearn is used to feed-back multiply the Θ opening degree command θ over the entire period of the FIRE mode, thereby correcting the SLD opening correction amount i / sld in the FIRE mode. Correction of the opening degree command Θ for each control cycle can be minimized. As a result, even if there is a large variation in the intake air amount due to structural factors, the temporal change pattern of the intake air amount deviates greatly from the standard opening command value θ0 pattern. Such a situation can be avoided, and stable operation of the internal combustion engine 1 can be performed without impairing the combustion state or emission state of the internal combustion engine 1.
[0451]
Further, in the present embodiment, during the FIRE mode operation, the operation of the internal combustion engine 1 or the start / run of the vehicle is performed by operating the accelerator pedal of the vehicle, and the operation other than the idling operation of the internal combustion engine 1 is performed. Even in the situation where the FIRE mode is interrupted, the bypass opening to the increase side of the intake air amount is manipulated (however, the intake air amount F / B control correction process is interrupted) and the estimated integrated intake air amount qair Continue calculating / pre and target integrated intake air quantity qair / cmd. Then, after the interruption operation is released, the intake air amount F / B control correction process is resumed, and the opening degree command Θ is corrected so that the estimated integrated intake air amount qair / pre converges to the target integrated intake air amount qair / cmd. To do. For this reason, during the operation of the FIRE mode, the operation of the internal combustion engine 1 other than the idling operation is performed, and even if the operation of interrupting the FIRE mode occurs, the catalyst device 3 within the FIRE mode limit time TFIRELMT Temperature rise and activation can be performed reliably.
[0452]
As for the ignition timing operation rotational speed F / B control, the adaptive sliding mode control (ignition timing side adaptive SLD control) is used for the feedback control processing as in the case of the intake air amount F / B control correction processing. Thus, the stability of the control of the internal combustion engine 1 to the target rotational speed ne / fire can be enhanced.
[0453]
In this case, in particular, the response designation characteristic of the ignition timing side adaptive SLD control, which is a response designation type control, is used, and the ignition timing command value iglog operated for the control of the rotational speed Ne is such that the value is on the retard side. Is set as a normal value of the pole pole / ig (= s4 / s3), which is a parameter for defining the damping speed, so as to slow down the damping speed of the rotational speed deviation En (= Ne-ne / fire). The ignition timing corresponding pole basic value pole / igtbl is variably determined. That is, generally, as the ignition timing is retarded, the change in the rotational speed Ne of the internal combustion engine 1 with respect to the change in the ignition timing becomes larger. In such a state, the decay rate of the rotational speed deviation En is increased. By setting the value of the pole pole / ig so as to be delayed, the change of the command value iglog determined by the ignition timing side adaptive SLD control is suppressed, and the rotation speed Ne is less than the target rotation speed ne / fire. Avoid sudden changes. On the contrary, on the advance side of the ignition timing, the value of the pole pole / ig is set so as to increase the decay speed of the rotational speed deviation En, so that the rotational speed Ne can quickly reach the target rotational speed ne / fire. To ensure proper followability. As a result, it is possible to improve stability while ensuring an appropriate speed response of control of the rotational speed Ne to the target rotational speed ne / fire.
[0454]
In the initial stage immediately after the start of the ignition timing operation rotational speed F / B control (until the rotational speed F / B elapsed time Δt / nfb reaches a predetermined value T / NFBX), the decay speed of the rotational speed deviation En is slowed down. As described above, the pole corresponding to the ignition timing corresponding pole basic value pole / igtbl is corrected by the time corresponding correction coefficient kigt to determine the pole pole / ig. As a result, it is possible to avoid a situation in which the rotational speed Ne rapidly changes immediately after the ignition timing operation rotational speed F / B control is started and the combustion state of the internal combustion engine 1 deteriorates.
[0455]
Further, in the present embodiment, when the ignition timing operation rotational speed F / B control is resumed after cancellation of the interruption operation in the FIRE mode, the rotational speed Ne of the internal combustion engine 1 becomes higher than the target rotational speed ne / fire. The pole pole / ig is determined by multiplying and correcting the ignition timing corresponding pole basic value pole / igtbl by the correction correction coefficient kignef corresponding to the rotation speed so that the damping speed of the rotation speed deviation En becomes slower as the distance is larger. Thus, when the ignition timing operation rotational speed F / B control is resumed, when the rotational speed Ne of the internal combustion engine 1 is far away from the target rotational speed ne / fire toward the high rotational speed side, a large rotational speed deviation En is obtained. Accordingly, it is possible to suppress the ignition timing command value iglog from suddenly changing to the retard side, and to ensure a stable operating state of the internal combustion engine 1.
[0456]
Furthermore, in the present embodiment, the absolute value of the pole pole / ig that defines the decay speed of the rotational speed deviation En related to the ignition timing operation rotational speed F / B control is set to be greater than the absolute value of the pole pole / ig. The absolute value of the pole pole / i that defines the decay speed of the intake deviation Eq is set to a larger value (see the processing in STEPs 17-7 and 17-8 in FIG. 17). In other words, the pole pole / ig is set to a value such that the decay rate of the rotational speed deviation En is faster than the decay rate of the intake air deviation Eq. Therefore, the feedback control for converging the rotational speed Ne of the internal combustion engine 1 to the target rotational speed ne / fire by the ignition timing operation rotational speed F / B control and the estimated integrated intake air amount qair / by the intake air amount F / B control correction processing. It is possible to stably control the rotational speed Ne of the internal combustion engine 1 to the target rotational speed ne / fire by avoiding mutual interference with feedback control for converging pre to the target integrated intake air amount qair / cmd.
[0457]
In the present embodiment, in the ignition timing side adaptive SLD control used for the ignition timing operation rotational speed F / B control, the internal combustion engine 1 is controlled from the ignition timing deviation command value DIG corresponding to the ignition timing command value iglog. As a system for generating the deviation speed DNE corresponding to the engine speed Ne, it is expressed by a discrete system model (rotation speed control target model), which makes the ignition timing side adaptive SLD control algorithm simple and suitable for computer processing. Can be.
[0458]
In particular, by modeling the control target of the ignition timing side adaptive SLD control in a discrete system, the switching function σ2 can be used without using the change speed of the rotational speed deviation En, as in the case of the intake side adaptive SLD control. It can be configured using time series data of only the rotational speed deviation En. As a result, the reliability of the value of the switching function .sigma.2 necessary for obtaining the ignition timing deviation command value DIG can be improved, and consequently the reliability of the ignition timing side adaptive SLD control can be improved.
[0459]
In the present embodiment, the bypass valve 7 is used as a flow control valve for increasing the intake air amount of the combustion chamber 4 of the internal combustion engine 1 in the FIRE mode, but by operating the opening of the throttle valve 5, The amount of intake air may be increased.
[0460]
Further, in the above-described embodiment, when the air-fuel ratio of the air-fuel mixture burned in the combustion chamber 4 is constant, the amount of heat generated by the internal combustion engine 1 (heat amount of exhaust gas) and thus the amount of heat given to the catalyst device 3 is the combustion chamber. Therefore, the integrated value of the intake air amount for each control cycle is used as the heat amount data representing the amount of heat given to the catalyst device 3. However, since the amount of heat generated by the internal combustion engine 1 slightly changes depending on the ignition timing of the internal combustion engine 1, when it is necessary to further improve the accuracy of the heat amount data representing the amount of heat given to the catalyst device 3, the instantaneous instantaneous (control The estimated value or the detected value of the intake air amount for each cycle) may be corrected according to the ignition timing at each time, and may be integrated to obtain the heat amount data. In addition, when it becomes necessary to change the air-fuel ratio of the air-fuel mixture, the estimated value or detected value of the intake air amount instantaneously (for each control cycle) is changed from time to time, as in the correction according to the ignition timing. The amount of heat data may be acquired by correcting according to the air-fuel ratio and integrating the corrected values. In any case, it is possible to cope with this by preparing a data table or the like in which a correction coefficient for performing correction according to the ignition timing and the air-fuel ratio is set in advance.
[0461]
It should be noted that the correction as described above when acquiring heat quantity data representing the heat quantity given to the catalyst device can be similarly applied when other parameters such as a fuel supply amount are used as the heat quantity data.
[Brief description of the drawings]
FIG. 1 is an overall configuration diagram of a control system for an internal combustion engine including an embodiment of the present invention.
FIG. 2 is a diagram schematically showing an intake system of an internal combustion engine of the system of FIG.
FIG. 3 is a diagram for explaining the basic operation of the system of FIG. 1;
FIG. 4 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 5 is a flowchart for explaining the operation of the system shown in FIG. 1;
FIG. 6 is a diagram for explaining the operation of the system of FIG. 1;
7 is a diagram for explaining the operation of the system of FIG. 1; FIG.
FIG. 8 is a diagram for explaining the operation of the system of FIG. 1;
FIG. 9 is a diagram for explaining the operation of the system of FIG. 1;
10 is a diagram for explaining the operation of the system of FIG. 1; FIG.
11 is a diagram for explaining the operation of the system of FIG. 1;
12 is a diagram for explaining the operation of the system of FIG. 1; FIG.
13 is a diagram for explaining the operation of the system of FIG. 1; FIG.
FIG. 14 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 15 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 16 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 17 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 18 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 19 is a diagram for explaining the operation of the system of FIG. 1;
20 is a diagram for explaining the operation of the system of FIG. 1;
FIG. 21 is a diagram for explaining the operation of the system of FIG. 1;
22 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 23 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 24 is a flowchart for explaining the operation of the system of FIG. 1;
FIG. 25 is a diagram for explaining the operation of the system of FIG. 1;
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Internal combustion engine, 3 ... Catalytic device, 7 ... Bypass valve (flow control valve), 8 ... Bypass passage (intake air passage), 25 ... Intake air amount operation means, 26 ... Ignition timing operation means

Claims (2)

During the idling operation in the operation mode for early activation of the catalytic device after the start of the internal combustion engine that releases the exhaust gas through the catalytic device, the intake air amount of the internal combustion engine is idling in a mode other than the operation mode. Intake air amount operation means for operating the opening of the flow control valve provided in the intake air passage of the internal combustion engine so that the intake air amount is increased by a predetermined amount from the intake air amount during operation , and the internal combustion engine during the increase of the intake air amount An internal combustion engine having an ignition timing operation means for operating the ignition timing of the internal combustion engine by feedback control so as to converge the engine rotation speed to a predetermined target rotation speed and correcting the ignition timing to the retard side. in the control unit, the ignition by the intake air amount increase and the ignition timing control means according to the intake air quantity control means of the operating mode for performing an early activation of the catalytic converter A period of operation, so as to run in the period from the start of said operating mode until the predetermined time has elapsed, the opening of pre-Symbol the flow control valve at the intake air amount increase, the more the atmospheric pressure low A control apparatus for an internal combustion engine, characterized in that the intake air amount operating means includes means for adjusting the opening according to the atmospheric pressure so as to increase the opening. The intake air amount operation means includes means for adjusting the opening degree of the flow rate control valve when the intake air amount is increased according to the atmospheric temperature so that the opening degree increases as the atmospheric temperature increases. The control apparatus for an internal combustion engine according to claim 1.

Images