JP3735506B2 - Body slip angle calculation device - Google Patents

Description

【００３２】
とする。ただし、ｃiはフィルタ伝達関数の係数、ｓはラプラス演算子である。
【００３３】
次に、伝達関数同定回路３４が依拠する演算式を導出しておく。なお、本実施の形態では、前処理フィルタ３２の演算を、伝達関数同定回路３４の演算に含めて実施する。
【００３４】
まず、同定すべき伝達関数は、路面外乱ΔＴdを加振入力として、このとき前処理フィルタ３２により検出された車輪速振動Δω1を応答出力とする２次のモデルとする。すなわち、
【００３５】
【数２】

【００３６】
の振動モデルを仮定する。ここに、ｖは車輪速度信号を観測するときに含まれる観測雑音である。（２）式を変形すると、次式を得る。
【００３７】
【数３】

【００３８】
まず、（３）式に（１）式の前処理フィルタを掛けて得られた式を離散化する。このとき、Δω1、ΔＴd、ｖは、サンプリング周期Ｔｓ毎にサンプリングされた離散化データΔω1 （ｋ）、ΔＴｄ （ｋ）、ｖ（ｋ）（ｋはサンプリング番号：ｋ＝１，２，３，．．．． ）として表される。また、ラプラス演算子ｓは、数４に示すように、所定の離散化手法を用いて離散化することができる。本実施の形態では、１例として、次の双一次変換により離散化するものとする。なお、ｄは１サンプル遅延演算子である。
【００３９】
【数４】

【００４０】
また、前処理フィルタの次数ｍは、２以上が望ましいので、本実施の形態では、演算時間も考慮してｍ＝２とし、これによって次式を得る。
【００４１】
【数５】

【００４２】
また、最小自乗法に基づいて、車輪速振動Δω1の各データから伝達関数を同定するために、（４）式を、同定すべきパラメータに関して一次関数の形式となるように、次式のように変形する。なお、”T”を行列の転置とする。
【００４３】
【数６】

【００４４】
である。上式において、θが同定すべき伝達関数のパラメータとなる。
【００４５】
次に、振動レベル演算回路１６は、車輪速度センサ１２によって検出された車輪速度信号のうち、少なくとも１つの共振点または少なくとも１つの反共振点を含む周波数が低周波数領域より大きい所定範囲の車輪速度信号を抽出するバンドパスフィルタ３８を備えている。バンドパスフィルタ３８には、バンドパスフィルタ３８より抽出された車輪速度信号に基づいて、車輪の振動の大きさを表す物理量、本実施の形態では、振動レベルを演算する演算回路４０が接続されている。
【００４６】
次に、タイヤグリップ状態検出回路２４を説明する。図３に示すように、タイヤグリップ状態検出回路２４は、車輪の路面に対するグリップ状態が高い状態である基準状態を検出する基準状態検出回路４２と、基準状態検出回路４２により検出された基準状態におけるμ勾配の平均値であるμ勾配初期値を記憶する路面μ勾配初期値記憶回路４４と、路面μ勾配初期値記憶回路４４に記憶されたμ勾配初期値と検出されたμ勾配値とを比較して、タイヤグリップ状態を表す値を出力する路面μ勾配値比較回路４６と、を備えている。
【００４７】
図４に示すように、基準状態検出回路４２は、車輪速度センサ１２により検出された左右輪の車輪速度、車速センサ２２により検出された車速、振動レベル演算回路１６により演算された振動レベル、ワイパ作動検出センサ２０により検出されたワイパ作動状態、及び温度センサ１８により検出さた環境温度が入力され、これらから車両が直進かつ一定速度で走行しかつ振動レベルが低い高μ路を走行していることを検出して信号を出力するＡＮＤ回路５２と、ＡＮＤ回路５２から信号が入力された場合、路面μ勾配演算回路１４からのμ勾配に出力を切り替える切替回路５４と、切替回路５４からの出力に基づいて車輪の路面に対するグリップ状態が高い状態である基準状態であることを示す高μ路フラグを出力する高μ路検出回路５６と、を備えている。
【００４８】
高μ路検出回路５６は、切替回路５４からの出力（路面μ勾配演算回路１４からのμ勾配）がしきい値α（予め実験等により定めた高μ路走行時のμ勾配の最大値に近い値（図８参照））より大きいか否かを判定する比較回路５８と、比較回路５８により路面μ勾配演算回路１４からのμ勾配がしきい値αより大きいと判定された場合には、判定された回数をカウント（変数ｉを１インクリメント）するカウンタ６０と、一定時間以上カウント値が更新されない場合にカウンタ６０のカウント値をリセット（ｉ＝０）するクロック回路６２と、カウンタ６０のカウント値（ｉ）が所定値Ｎより大きいか否かを判定する比較器６４と、比較器６４によりカウンタ６０のカウント値（ｉ）が所定値Ｎより大きいと判定された場合に、メモリ６６に記憶された高μ路フラグを出力し、カウント値（ｉ）が所定値Ｎより大きいと判定されなかった場合に、メモリ６８に記憶された非高μ路フラグを出力するように切り替える切替器７０と、を備えている。なお、カウンタ６０のカウント値（ｉ）が所定値Ｎより大きいと判定された場合に、メモリ６６に記憶された高μ路フラグを出力するようにしているのは、高μ路でなくてもノイズや路面状態等によりμ勾配がしきい値より大きくなる場合があり、これにより高μ路フラグが出力されないようにするためである。
【００４９】
図５に示すように、路面μ勾配初期値記憶回路４４は、高μ路フラグ及び車速センサ２２からの信号（走行開始）が入力された場合、即ち、車両が走行を開始し、車輪の路面に対するグリップ状態が高い状態の場合に信号を出力するＡＮＤ回路７４と、ＡＮＤ回路７４からの信号が入力されたときからの所定時間内のμ勾配を平均するμ勾配平均値算出回路７２と、μ勾配平均値算出回路７２により算出されたμ勾配平均値をμ勾配初期値μＩとして記憶するμ勾配初期値記憶回路７６と、を備えている。
【００５０】
図６に示すように、路面μ勾配値比較回路４６は、路面μ勾配演算回路１４により演算されたμ勾配の平均値μｒを移動平均処理等により算出するμ勾配平均値算出回路７８と、図示しないタイマからのタイマ信号が入力される毎にオンする切替器８０と、μ勾配初期値μＩとμ勾配の平均値μｒとの比（μｒ／μＩ）を算出してタイヤグリップ状態を表す値として出力するμ勾配値比較回路８２と、を備えている。
【００５１】
次に、本実施の形態の作用を説明する。本実施の形態に係る車体スリップ角演算回路３０は、複数のグリップ状態各々毎に対応して予め定められかつ上記車速センサ２２により検出された車速、ヨーレートセンサ２６により検出されたヨーレート、及び舵角センサ２８により検出された舵角に基づいて車体スリップ角を演算する複数の演算処理の内、上記タイヤグリップ状態検出回路２４により検出されたグリップ状態に対応する演算処理を選択的に実行して車体スリップ角を演算するものである。
【００５２】
まず、タイヤグリップ状態検出回路２４によるグリップ状態の検出内容を説明する。このために、路面μ勾配演算回路１４の演算内容及び振動レベル演算回路１６の演算内容を最初に説明する。
【００５３】
伝達関数同定回路３４では、検出された車輪速振動Δω1の離散化データを(１１)式に順次当てはめた各データに対し、最小自乗法を適用することによって、未知パラメータθを推定し、これにより伝達関数を同定する。
【００５４】
具体的には、検出された車輪速振動Δω1を離散化データΔω（ｋ）（ｋ＝１，２，３，．．．）に変換し、該データをＮ点サンプルし、次式の最小自乗法の演算式を用いて、伝達関数のパラメータθを推定する。
【００５５】
【数７】

【００５６】
ここに、記号”＾”の冠した量をその推定値と定義することにする。
【００５７】
また、上記最小自乗法は、次の漸化式によってパラメータθを求める逐次型最小自乗法として演算してもよい。
【００５８】
【数８】

【００５９】
ここに、ρは、いわゆる忘却係数で、通常は０．９５〜０．９９の値に設定する。このとき、初期値は、
【００６０】
【数９】

【００６１】
とすればよい。
【００６２】
また、上記最小自乗法の推定誤差を低減する方法として、種々の修正最小自乗法を用いてもよい。本実施の形態では、補助変数を導入した最小自乗法である補助変数法を用いた例を説明する。該方法によれば、（９）式の関係が得られた段階でｍ（ｋ）を補助変数として、次式を用いて伝達関数のパラメータを推定する。
【００６３】
【数１０】

【００６４】
また、逐次演算は、以下のようになる。
【００６５】
【数１１】

【００６６】
補助変数法の原理は、以下の通りである。（１６）式に（１５）式を代入すると、
【００６７】
【数１２】

【００６８】
となるので、（１９）式の右辺第２項が零となるように補助変数を選べばθの推定値は、θの真値に一致する。そこで、本実施の形態では、補助変数として、ζ（ｋ）＝［−ξy1（ｋ）−ξy2（ｋ）］Tを式誤差ｒ（ｋ）と相関を持たないほどに遅らせたものを利用する。すなわち、
ｍ（ｋ）＝［−ξy1（ｋ−Ｌ）−ξy2（ｋ−Ｌ）］T （２０）
とする。ただし、Ｌは遅延時間である。
【００６９】
上記のようにして伝達関数を同定した後、演算回路３６において、路面μ勾配Ｄ０に関係する物理量を、
【００７０】
【数１３】

【００７１】
と演算する。このように（２１）式により路面μ勾配Ｄ０に関係する物理量を演算できると、例えば、該物理量が小さいとき、タイヤと路面との間の摩擦特性が飽和状態であると容易に判定できる。
【００７２】
ところで、車輪速度センサ１２により、検出された車輪速度信号は、周波数に応じて分解すると、図９に示すように、２つの共振点と１つの反共振点を有する。２つの共振点のうち周波数が小さい側の共振点は、タイヤ慣性などに基づく前後共振点であり、周波数はｆ1（１５〜２０）Ｈｚである。また、２つの共振点のうち周波数が高い側の共振点は、タイヤの空気圧やタイヤゴム弾性に基づく捩れ共振点であり、周波数はｆ3（３５〜４０）Ｈｚである。そして、車輪速度信号には、色々な信号に対して不感帯の反共振点を有し、周波数ｆ2（２０〜２５）Ｈｚである。本実施の形態に係るバンドパスフィルタ３８は、車輪速度センサ１２により検出された車輪速度信号のうち、捩れ共振点（周波数ｆ3）含む所定範囲Δｆの車輪速度信号を抽出している。なお、バンドパスフィルタ３８は、捩れ共振点以外の、前後共振点や反共振点を含む所定範囲Δｆの車輪速度信号を抽出するようにしてもよい。
【００７３】
演算回路４０は次式より定まる車輪速度信号の振動レベルＧ（Ｎ）を演算する。なお、バンドパスフィルタ３８の出力はω（ｋ）である。
【００７４】
【数１４】

【００７５】
なお、演算回路４０は、実際には、演算タイミング毎に、
【００７６】
【数１５】

【００７７】
の漸化式を逐次的に演算する。このように、演算回路４０で振動レベルを演算するのは、路面μ勾配は、突起乗り越し等により、適正に推定されないからである。
【００７８】
一方、このようにして得られた振動レベルと路面μ勾配の推定値とを、アスファルト路面、低μ路路面においてプロットすると、図１０に示すようにアスファルト路面と低μ路路面とでは明確に区別されて認識されるとともに、アスファルト路面中に突起を乗り越したとしても、低μ路における領域と明確に区別されて認識される。上記事実に鑑み本発明者は、種々の路面状態毎に振動レベルと路面μ勾配との推定値との関係を求めたところ、図１１に示すように、振動レベルと路面μ勾配の推定値とは各路面状態、例えば低μ路、アスファルト路、石畳路、圧雪路、砂利・非整地路毎に領域毎に明確に区別されて認識出来ることが実験の結果得られた。即ち、例えば、圧雪路は路面μ勾配の推定値としては高μ路（アスファルト路、石畳路）より若干小さい値となっているが、振動レベルの違いによって領域に差が表れている。砂利道等は更に振動レベルが高い領域にあるが、路面μ勾配の推定値は高μ路より低い値となっている。
【００７９】
ここで、前述したように、基準状態検出回路４２のＡＮＤ回路５２には、車輪速度センサ１２により検出された左右輪の車輪速度、車速センサ２２により検出された車速、振動レベル演算回路１６により演算された振動レベル、ワイパ作動検出センサ２０により検出されたワイパ作動状態、及び温度センサ１８により検出さた環境温度が入力される。
【００８０】
即ち、車輪速度センサ１２により検出された左右輪の車輪速度差から車両が直進しているのか判断することができる。車速センサ２２により検出された車速から車両が一定速度で走行しているか判断することができる。ワイパ作動検出センサ２０により検出されたワイパ作動状態から、ワイパが作動していなければ雨が降っていない、即ち、ＤＲＹ路面であると判断することができる。そして、温度センサ１８により検出さた環境温度で、路面が凍結していないか判断することができる。そして、振動レベル演算回路１６により演算された振動レベルと温度センサ１８により検出さた環境温度とから、振動レベルが小さく（例えば０．０１など）かつ路面凍結していなければ、図１１に示した関係から、アスファルト路面（高μ路）を走行していると判断することができる。そして、ＡＮＤ回路５２は、車両が直進かつ一定速度で走行しかつ振動レベルが低い高μ路を走行していることを検出すると信号を出力する。
【００８１】
切替回路５４は、ＡＮＤ回路５２から信号が入力された場合、路面μ勾配演算回路１４からのμ勾配に出力を切り替える。比較器５８は、切替回路５４からの出力（路面μ勾配演算回路１４からのμ勾配）がしきい値αより大きいか否かを判定する。カウンタ６０は、比較回路５８により路面μ勾配演算回路１４からのμ勾配がしきい値αより大きいと判定された場合には、判定された回数をカウント（変数ｉを１インクリメント）する。なお、一定時間以上カウント値が更新されない場合にカウンタ６０のカウント値はクロック回路６２からの信号によりリセット（ｉ＝０）される。比較器６４は、カウンタ６０のカウント値（ｉ）が所定値Ｎより大きいか否かを判定する。そして、切替器７０は、比較器６４によりカウンタ６０のカウント値（ｉ）が所定値Ｎより大きいと判定された場合に、メモリ６６に記憶された高μ路フラグを出力する。なお、カウンタ６０のカウント値（ｉ）が所定値Ｎより大きいと判定された場合に、メモリ６６に記憶された高μ路フラグを出力しているので、高μ路でなくてもノイズや路面状態等によりμ勾配がしきい値より大きくなる場合があっても、これにより高μ路フラグが出力されないようにすることができる。
【００８２】
路面μ勾配初期値記憶回路４４のＡＮＤ回路７４は、高μ路フラグ及び車速センサ２２からの信号（走行開始）が入力された場合、即ち、車両が走行を開始し、車輪の路面に対するグリップ状態が高い状態の場合に信号を出力する。μ勾配平均値算出回路７２は、ＡＮＤ回路７４からの信号が入力されたときからの所定時間内のμ勾配を平均する。そして、μ勾配初期値記憶回路７６は、μ勾配平均値算出回路７２により算出されたμ勾配平均値をμ勾配初期値μＩとして記憶する。
【００８３】
一方、路面μ勾配値比較回路４６のμ勾配平均値算出回路７８は、路面μ勾配演算回路１４により演算されたμ勾配の平均値μｒを移動平均処理等により算出する。そして、上記タイマ信号が入力されると切替器８０はオンし、これにより、μ勾配値比較回路８２にμ勾配平均値算出回路７８により演算されたμ勾配の平均値μｒが入力され、μ勾配初期値μＩとμ勾配の平均値μｒとの比（μｒ／μＩ）が算出されてタイヤグリップ状態を表す値として出力される。
【００８４】
次に、本実施の形態に係る車体スリップ角演算回路３０の具体的作用を、図７に示したフローチャートを参照して説明する。
【００８５】
ステップ１０２で、各車輪毎に配置された上記μ勾配値比較回路８２からの出力に基づいて、４輪のグリップレベルが第１のしきい値以上か否かを判断する。なお、この第１のしきい値は、車輪の横力に対して車体スリップ角が線形である線形状態であるか否かを判断するための値であり、予め定められている。即ち、４輪のグリップレベルが第１のしきい値以上の場合には、車輪の横力に対して車体スリップ角が線形である線形状態であると判断することができる。
【００８６】
４輪のグリップレベルが第１のしきい値以上でないと判断された場合には、ステップ１０４で、４輪のグリップレベルが第２のしきい値以下か否かを判断する。なお、この第２のしきい値は、車両の旋回が限界である旋回限界状態であるか否かを判断するための値であり、予め定められている。即ち、４輪のグリップレベルが第２のしきい値以下の場合には、旋回限界状態であると判断することができる。ステップ１０４で否定判定された場合は、車輪の横力に対して車体スリップ角が線形である線形状態でなくかつ旋回限界状態でもない中間の状態である。なお、ステップ１０４は、請求項４における旋回限界状態検出手段に対応する。
【００８７】
車輪の横力に対して車体スリップ角が線形である線形状態の場合には、ステップ１０６で、第１のスリップ角推定処理（本発明の演算処理、詳細は後述する）を行う。一方、車輪の横力に対して車体スリップ角が線形である線形状態でなくかつ旋回限界状態でもない中間の状態の場合には、ステップ１０８で、第２のスリップ角推定処理（本発明の演算処理、詳細は後述する）を行う。そして、車両の旋回が限界である旋回限界状態の場合には、ステップ１１０、１１２で、擬似積分（本発明の演算処理、詳細は後述する）を行う。
【００８８】
次に、上記第１のスリップ角推定処理を説明する。車輌のヨーレートの検出値をγとし、車体のスリップ角の推定値βh に対するヨーレートの検出値γと推定値γh との偏差（γ−γh ）のフィードバックゲインをＫb とし、ヨーレートの推定値γh に対するヨーレートの検出値γと推定値γh との偏差（γ−γh ）のフィードバックゲインをＫg とし、車輌の重量をｍとし、車速をＶとし、左右前輪及び左右後輪の横力をそれぞれＦf 及びＦr とし、車輌の重心より前輪及び後輪までの車輌前後方向の距離をそれぞれＬf 及びＬr とすると、車輌のスリップ角の推定値βh の微分値βhd及び車輌の推定ヨーレートγh の微分値γhdはそれぞれ下記の数１６及び数１７の基本方程式により表される。
【数１６】

【００８９】
【数１７】

また前輪及び後輪のスリップ角をそれぞれβf 及びβr とし、前輪及び後輪のコーナリングパワーをそれぞれＣf 及びＣr とすると、車輌モデルの線形理論によれば、前輪及び後輪の横力Ｆf 及びＦr はそれぞれ下記の数１８及び数１９の式により表される。
【数１８】

【００９０】
【数１９】

【００９１】
周知の如く、コーナリングパワーＣf 及びＣr は前輪及び後輪のスリップ角βf 及びβr が比較的小さい場合にのみ一定であり、スリップ角βf 及びβr が増大するにつれて漸次減少する。前輪及び後輪のスリップ角βf 及びβr の推定値βfh及びβrhは前輪の舵角δf （δfr及びδflの平均値であってよい）、後輪の舵角δr （δrr及びδrlの平均値であってよい）、車体のスリップ角の推定値βh 、車輌の重心より前輪及び後輪までの長手方向の距離Ｌf 及びＬr 、ヨーレートの推定値γh 、車速Ｖに基づき推定可能であり、それぞれ下記の数２０及び数２１の式により表される。
【００９２】
【数２０】

【００９３】
【数２１】

【００９４】
上記数２０及び数２１の推定値βfh及びβrhをそれぞれ数１８及び数１９のβf 及びβr に代入し、数１６及び数１７の基本方程式に数１８及び数１９を代入し、上記基本方程式を行列形式にて表すと下記の数２２の如くになる。
【００９５】
【数２２】

【００９６】
上記数２２を下記の数２３の如く表現することができる。
【００９７】
【数２３】

【００９８】
数２３を更に下記の数２４の通り書き直す。
【００９９】
【数２４】

従って（Ａ−ＬＣ）^-1（−Ｂ＊δ−γ＊Ｋ）をＸ_Cとおくと、数２４を下記の数２５の如く書き直すことができる。
【０１００】
【数２５】

【０１０１】
ＸがＸc に収束するよう数２５を解くことができる。この系は二次の系であるので、二つの極ｐ1 及びｐ2 を有する。従ってβc 及びγc はこれらの極ｐ1 及びｐ2 を用いてそれぞれ下記の数２６及び数２７にて表される。
【０１０２】
【数２６】

【０１０３】
【数２７】

【０１０４】
尚上記数２６及び数２７に於けるβo 及びγo は下記の数２８の如く定義される。
【０１０５】
【数２８】

【０１０６】
またフィードバックゲインＫb 及びＫg は上記二つの極ｐ1 及びｐ2 を用いてそれぞれ下記の数２９及び数３０の如く表される。尚上記数２２に示されている如く、ａ11、ａ12、ａ21、ａ22は行列Ａの成分である。
【０１０７】
【数２９】

【０１０８】
【数３０】

【０１０９】
車輪のスリップ角βf 及びβr が比較的小さいときには、フィードバックゲインＫb 及びＫg は、その第一のフィードバックゲイン成分Ｋb1及びＫg1として、本願出願人の出願にかかる特願平７−１０７８２号の明細書に記載されている如く求められ、この場合Ｃ11、Ｃ12、Ｃ21、Ｃ22を正の定数として極ｐ1 及びｐ2はそれぞれ下記の数３１及び数３２にて表される。
【０１１０】
【数３１】

【０１１１】
【数３２】

【０１１２】
次に、第２のスリップ角推定処理を説明する。車輌がスピンするほどに車輪のスリップ状態が進行すると、車体のスリップ角及びヨーレートの何れも増大するが、スリップ角はヨーレートとは逆の符号にて増大する。従って数１１の関係に於いてヨーレートγの比例係数が負になるよう、換言すればスリップ角の推定値がヨーレートの推定値の増大方向とは逆に増大するよう設定されれば、実際の車輌の状況と車輌モデルとのずれが大きくなった場合にも推定誤差が小さく抑えられ、これにより車輌のスリップ角の推定精度が向上する。またヨーレートの推定値はヨーレートの検出値に収束しなければならないので、数２７に於けるヨーレートγの比例係数は正でなければならない。従って非線形領域に於いてはｐ1 及びｐ2 は下記の数３３及び数３４の条件を満たすよう設定される。
【０１１３】
【数３３】

【０１１４】
【数３４】

【０１１５】
以上の二つの条件を満たす極ｐ1 及びｐ2 の値が図１２及び図１３に於いてハッチングにて示された範囲の値として図示されている。特に図１２は車輌の低速走行時について極ｐ1 及びｐ2 の範囲を示し、図１３は車輌の高速走行時について極ｐ1 及びｐ2 の範囲を示している。極ｐ1 及びｐ2 の範囲は車速の増大につれて図１２の範囲より図１３の範囲に漸次変化する。
【０１１６】
車輌モデルの非線形理論によれば、前輪及び後輪の横力Ｆf 及びＦr は図１４に示されたグラフにより表される。前輪及び後輪のスリップ角βf 及びβr が図１４に於いて例えばβm にて示された値の近傍にまで増大し、前輪及び後輪のスリップ角βf 及びβr とそれらの車輪の横力Ｆf 及びＦr との間の比例関係が大きく失われると、車体のスリップ角の推定は更に修正される。
【０１１７】
コーナリングパワーＣf 及びＣr は時々刻々変化するが、前輪及び後輪の横力Ｆf 及びＦr を便宜的に下記の数２０及び数２１により表現することができる。尚下記の数３５及び数３６に於いて、ΔＦf 及びΔＦr はコーナリングパワーＣp の誤差や路面の摩擦係数μの推定誤差等を含む推定誤差である。またＦf0及びＦr0は定数ではなくそれぞれβfh及びβrhの関数になるが、βfh及びβrhの現在値βm の近傍のみを考えるとすればこれらを定数に近似することができる。
【０１１８】
【数３５】

【０１１９】
【数３６】

【０１２０】
数３５及び数３６のＦf 及びＦr を上記数３１及び数３２に代入し、その結果を行列形式にて表現すれば下記の数３７の如くになる。
【０１２１】
【数３７】

【０１２２】
数３７をラプラス変換すると下記の数３８の如く表される。
【０１２３】
【数３８】

【０１２４】
タイヤ横力の非線形領域（但し、上記中間状態）に於いて推定される横力は大きい誤差を含んでいるので、フィードバック係数αが調節されることにより、前輪のタイヤグリップが飽和状態にあるときにはΔＦf よりβh への伝達関数が０にされ、後輪のタイヤグリップが飽和状態にあるときにはΔＦr よりβh への伝達関数が０にされ、これにより横力の推定誤差による車輌のスリップ角βの推定誤差が低減される。
【０１２５】
ΔＦf よりβh への伝達関数を０にするためのフィードバック係数αは下記の数３９にて表され、ΔＦr よりβh への伝達関数を０にするためのフィードバック係数αは下記の数４０にて表される。
【０１２６】
【数３９】

【０１２７】
【数４０】

【０１２８】
次に、上記擬似積分を説明する。擬似積分は、旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算する誤差減少演算処理であり、
車体スリップ角を前回演算するのに用いた物理量を無視して、車体スリップ角を今回演算するために新たに検出された前記物理量を積分することにより、前記旋回限界状態であることに起因する誤差の影響を減少させるものである。
【０１２９】
即ち、ステップ１１０で、検出された車速から近似した前後方向速度Ｖｘ、車両に作用する横方向加速度Ｇｙ、及び検出されたヨーレートγの物理量から、ｄβ＝Ｇｙ／Ｖｘ−γを演算し、ステップ１１２で、βest＝Ｔ／（１＋Ｔ*ｓ ）ｄβを擬似積分する。即ち、車体スリップ角を前回演算するのに用いた物理量を無視して、車体スリップ角を今回演算するために新たに本ステップ１１２が開始されたときから例えば２０［μＳ］経過するまでの間に検出された前記物理量を積分する。なお、ｓは、コンピュータに実装するため、ラプラス演算子で式（４）を用いて離散化する。
【０１３０】
なお、第１のスリップ角推定処理後、ステップ１１４で、第１のスリップ角推定処理により得られたスリップ角β１が０以上の場合には、スリップ角β１として、スリップ角β１と所定値Ａ０との内で小さい値をとり、第１のスリップ角推定処理により得られたスリップ角β１が０より小さい場合には、スリップ角β１として、スリップ角β１と所定値Ａ０の負の値との内で大きい値をとる。
【０１３１】
以上説明したように本実施の形態は、車輪の路面に対する複数のグリップ状態各々毎に対応して、車体スリップ角を演算する複数（３つ）の演算処理を予め定め、検出されたグリップ状態に対応する演算処理を選択的に実行するので、車輪の路面に対するグリップ状態に係わらず車体スリップ角を正確に演算することができる。
【０１３２】
また、旋回限界状態が検出された場合には、旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算する誤差減少演算処理を実行するので、誤差の影響が時に大きい旋回限界状態の場合においても、車体スリップ角を正確に演算することができる。なお、この場合、車体スリップ角を前回演算するのに用いた物理量を無視して、車体スリップ角を今回演算するために新たに検出された物理量を積分することにより、旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算するので、上記物理量に含まれかつ旋回限界状態による過去の誤差の影響を無くすことができ、誤差の影響が時に大きい旋回限界状態の場合においても、車体スリップ角を正確に演算することができる。
【０１３３】
そして、車輪の横力に対して車体スリップ角が線形である線形状態が検出されかつ演算された車体スリップ角が限定値を超えた場合は、演算された車体スリップ角を限定値に規制するので、検出精度の悪化を防止することができる。
【０１３４】
以上説明した実施の形態では、タイヤグリップ状態（旋回限界状態）を路面μ勾配を用いて算出しているが、本発明はこれに限定されるものではなく、路面μ勾配と等価なスリップ速度に対する制動トルクの勾配である制動トルク勾配、及び駆動トルクの勾配である駆動トルク勾配などを用いて同様に算出するようにしてもよい。即ち、タイヤグリップ状態を、路面μ勾配、制動トルク勾配、及び駆動トルク勾配等の車輪のすべり易さを表す物理量に基づいて算出することができる。
【０１３５】
【発明の効果】
以上説明したように本発明は、車輪の路面に対する複数のグリップ状態各々毎に対応して、車体スリップ角を演算する複数の演算処理を予め定め、検出されたグリップ状態に対応する演算処理を選択的に実行するので、車輪の路面に対するグリップ状態に係わらず車体スリップ角を正確に演算することができる、という効果を有する。
【０１３６】
請求項２記載の発明は、旋回限界状態が検出された場合に、物理量検出手段により検出された上記物理量に含まれかつ旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算する誤差減少演算処理を実行するので、誤差の影響が時に大きい旋回限界状態の場合においても、車体スリップ角を正確に演算することができる、という効果を有する。
【０１３７】
請求項３記載の発明は、車体スリップ角を前回演算するのに用いた物理量を無視して、車体スリップ角を今回演算するために新たに検出された物理量を積分することにより、旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算するので、上記物理量に含まれかつ旋回限界状態による過去の誤差の影響を無くすことができ、誤差の影響が時に大きい旋回限界状態の場合においても、車体スリップ角を正確に演算することができる、という効果を有する。
【０１３８】
請求項４記載の発明は、車輪の横力に対して車体スリップ角が線形である線形状態が検出されかつ演算された車体スリップ角が限定値を超えた場合は、演算された車体スリップ角を限定値に規制するので、検出精度の悪化を防止することができる、という効果を有する。
【０１３９】
請求項５記載の発明は、旋回限界状態が検出された場合に、物理量検出手段により検出された上記物理量に含まれかつ旋回限界状態であることに起因する誤差の影響を減少させた車体スリップ角を演算するので、誤差の影響が時に大きい旋回限界状態の場合においても、車体スリップ角を正確に演算することができる、という効果を有する。
【図面の簡単な説明】
【図１】本発明の第１の実施の形態に係る車体スリップ演算回路のブロック図である。
【図２】路面μ勾配演算回路及び振動レベル演算回路のブロック図である。
【図３】タイヤグリップ状態検出回路のブロック図である。
【図４】基準状態検出回路を示す図である。
【図５】路面μ勾配初期値記憶回路を示す図である。
【図６】路面μ勾配値比較回路を示す図である。
【図７】本実施の形態の作用を示すフローチャートである。
【図８】基準状態検出回路内におけるしきい値を説明する説明図である。
【図９】車輪速度信号の周波数と振幅との関係を示したグラフである。
【図１０】車輪速度信号の振動レベルと路面μ勾配との、アスファルト路及び低μ路の関係を示したグラフである。
【図１１】車輪速度信号の振動レベルと路面μ勾配との路面状態毎の関係を示したグラフである。
【図１２】車輌の低速走行時について第二のフィードバックゲイン演算用パラメータｐ1及びｐ2 の適正範囲を示すグラフである。
【図１３】車輌の高速走行時について第二のフィードバックゲイン演算用パラメータｂ1及びｐ2 の適正範囲を示すグラフである。
【図１４】前輪及び後輪のスリップ角βf 及びβr と横力Ｆf 及びＦr との間の関係を示すグラフである。
【符号の説明】
２４ タイヤグリップ状態検出回路（グリップ状態検出手段、旋回限界状態検出手段）
２２ 車速センサ（物理量検出手段）
２６ ヨーレートセンサ（物理量検出手段）
２８ 舵角センサ（物理量検出手段）
３０ 車体スリップ角演算回路（演算実行手段、車体スリップ角演算手段）[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a vehicle body slip angle calculation device, and more particularly to a vehicle body slip angle calculation device that calculates a vehicle body slip angle based on a physical quantity representing the behavior of the vehicle body.
[0002]
[Prior art]
Conventionally, as described in JP-A-9-58444, the slip angle β of the vehicle body can be obtained by dividing the lateral speed Vy of the vehicle by the longitudinal speed Vx of the vehicle. The lateral velocity Vy can be obtained by integrating the translational acceleration acting on the vehicle. Further, the translational acceleration can be expressed as Gy−Vx · γ using the longitudinal velocity Vx, the lateral acceleration Gy acting on the vehicle, and the yaw rate γ around the center of gravity C. Therefore, the slip angle β of the vehicle body can be expressed as β = ∫ {(Gy / Vx) −γ} dt. Gy and γ can be detected by a lateral acceleration sensor and a yaw rate sensor, respectively. Vx can be approximated by the vehicle speed V. For this reason, the slip angle β of the vehicle body can be easily calculated in a vehicle equipped with a lateral acceleration sensor, a yaw rate sensor, and a vehicle speed sensor.
[0003]
[Problems to be solved by the invention]
However, when the grip state of the wheel with respect to the road surface changes, that is, for example, when the crossing gradient of the road surface on which the vehicle travels increases, the crossing gradient component is included in the lateral acceleration Gy, and the vehicle slip angle is accurately determined. Can not ask for.
[0004]
The present invention has been made in view of the above-described facts, and an object of the present invention is to propose a vehicle body slip angle calculation device that can accurately calculate the vehicle body slip angle regardless of the grip state of the wheels with respect to the road surface.
[0005]
[Means for Solving the Problems]
In order to achieve the above object, the invention according to claim 1 provides a grip state for the road surface of the wheel. Using the road surface μ gradient Grip state detection means for detecting, physical quantity detection means for detecting a physical quantity indicating the behavior of the vehicle body necessary for calculating the vehicle body slip angle, and predetermined by each grip state and detected by the physical quantity detection means Based on the physical quantity The first slip angle estimation process for estimating the slip angle of the vehicle body using the lateral force of the front and rear wheels in the vehicle equation of motion as a linear characteristic and the lateral force of the front and rear wheels in the vehicle equation of motion as a nonlinear characteristic. Including a second slip angle estimation process for estimating a slip angle Calculation execution means for selectively executing calculation processing corresponding to the grip state detected by the grip state detection means among a plurality of calculation processing for calculating the vehicle body slip angle is provided.
[0006]
The grip state detection means detects the grip state of the wheel with respect to the road surface, and the physical quantity detection means detects a physical quantity representing the behavior of the vehicle body necessary for calculating the vehicle body slip angle.
[0007]
The calculation execution means predetermines a plurality of calculation processes for calculating the vehicle body slip angle based on the physical quantity detected by the physical quantity detection means, corresponding to each of the plurality of grip states. The calculation execution means selectively executes calculation processing corresponding to the grip state detected by the grip state detection means.
[0008]
As described above, since a plurality of calculation processes for calculating the vehicle body slip angle corresponding to each of the plurality of grip states for the road surface of the wheel are determined in advance, and the calculation process corresponding to the detected grip state is selectively executed. The vehicle body slip angle can be accurately calculated regardless of the grip state of the wheel with respect to the road surface.
[0009]
Here, as in the invention described in claim 2, the grip state includes a turning limit state in which turning of the vehicle is a limit, and the calculation execution means includes: Including an error reduction calculation process for calculating a vehicle body slip angle that is included in the physical quantity detected by the physical quantity detection means and reduces the influence of the error caused by being in a turning limit state; When the turning limit state is detected by the grip state detection means , Wrong Difference reduction calculation processing is executed.
[0010]
Thus, when a turning limit state is detected, an error for calculating a vehicle body slip angle that is included in the physical quantity detected by the physical quantity detection means and that reduces the influence of the error caused by the turning limit state. Since the reduction calculation process is executed, the vehicle body slip angle can be accurately calculated even in the turning limit state where the influence of the error is sometimes great.
[0011]
In this way, when executing the error reduction calculation process, the calculation execution means ignores the physical quantity used for the previous calculation of the vehicle body slip angle as in claim 3 and sets the vehicle body slip angle this time. By integrating the newly detected physical quantity for calculation, a vehicle body slip angle is calculated in which the influence of the error due to the turning limit state is reduced.
[0012]
In this way, ignoring the physical quantity used to calculate the vehicle body slip angle last time and integrating the newly detected physical quantity to calculate the vehicle body slip angle this time, it is caused by the turning limit state. The vehicle body slip angle with reduced influence of error is calculated, so it is possible to eliminate the influence of past errors included in the physical quantity and due to the turning limit state, even in the case of turning limit state where the influence of error is sometimes large The vehicle body slip angle can be accurately calculated.
[0013]
On the other hand, the grip state includes a linear state in which the vehicle body slip angle is linear with respect to the lateral force of the wheel, and the calculation execution means detects the linear state by the grip state detection means. The slip angle is calculated and calculated by the first slip angle estimation process. When the calculated vehicle body slip angle exceeds the limit value, the calculated vehicle body slip angle is restricted to the limit value.
[0014]
As described above, when a linear state in which the vehicle body slip angle is linear with respect to the lateral force of the wheel is detected and the calculated vehicle body slip angle exceeds the limit value, the calculated vehicle body slip angle is restricted to the limit value. Therefore, it is possible to prevent deterioration in detection accuracy.
[0015]
The invention according to claim 5 provides a turning limit state in which turning of the vehicle is a limit. Using the road surface μ gradient When the turning limit state is detected by the turning limit state detecting means to detect, the physical quantity detecting means for detecting the physical quantity indicating the behavior of the vehicle body necessary for calculating the vehicle body slip angle, and the turning limit state detecting means, Based on the physical quantity detected by the physical quantity detection means, a vehicle body slip that is included in the physical quantity and that reduces the influence of errors caused by the turning limit state Corner Vehicle body slip angle calculating means for calculating.
[0016]
The turning limit state detecting means detects a turning limit state in which turning of the vehicle is a limit. The physical quantity detection means detects a physical quantity representing the behavior of the vehicle body necessary for calculating the vehicle body slip angle.
[0017]
When the turning limit state detecting means detects the turning limit state, the vehicle body slip angle calculating means is based on the physical quantity detected by the physical quantity detecting means and includes an error caused by being in the turning limit state. Body slip with reduced impact Corner Calculate.
[0018]
As described above, when the turning limit state is detected, the vehicle body slip angle included in the physical quantity detected by the physical quantity detection means and reducing the influence of the error caused by the turning limit state is calculated. Even in the case of a turning limit state where the influence of the error is sometimes great, the vehicle body slip angle can be calculated accurately.
[0019]
Note that, as described above, the vehicle body slip angle calculation means in the present invention ignores the physical quantity used for the previous calculation of the vehicle body slip angle, and the physical quantity newly detected to calculate the vehicle body slip angle this time. May be calculated to reduce the influence of the error caused by the turning limit state. As in the above, it is possible to eliminate the influence of past errors included in the physical quantity and due to the turning limit state, and the vehicle body slip angle is accurately calculated even in the turning limit state where the influence of the error is sometimes large. be able to.
[0020]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0021]
As shown in FIG. 1, the vehicle body slip angle calculation device according to the present embodiment is based on the wheel speed detected by the wheel speed sensor 12 with the gradient of the friction coefficient μ between the wheel and the road surface with respect to the slip speed. Based on a road surface μ gradient calculation circuit 14 for calculating a gradient of a road surface μ (hereinafter referred to as μ gradient) and a wheel speed detected by the wheel speed sensor 12, a vibration level which is a physical quantity indicating the magnitude of vibration of the wheel is obtained. And a vibration level calculation circuit 16 for calculation.
[0022]
In addition, the vehicle body slip angle calculation device according to the present embodiment includes the environmental temperature detected by the temperature sensor 18, the operating state of the wiper detected by the wiper operation detection sensor 20, the vehicle speed detected by the vehicle speed sensor 22, and the road surface μ. Tire grip state detection as a grip state detection means of the present invention that detects the grip state of the wheel against the road surface based on the μ gradient calculated by the gradient calculation circuit 14 and the vibration level calculated by the vibration level calculation circuit 16. A circuit 24 is provided.
[0023]
Further, the vehicle body slip angle calculation device according to the present embodiment is predetermined for each of a plurality of (three (but not limited to three)) grip states in the present embodiment and the vehicle speed sensor 22 described above. The vehicle body slip angle is calculated based on the vehicle speed detected by the yaw rate, the yaw rate detected by the yaw rate sensor 26, a sensor (not shown) for detecting lateral acceleration acting on the vehicle, and the rudder angle detected by the rudder angle sensor 28. Calculation execution means and vehicle slip angle calculation means for selectively calculating a vehicle body slip angle by selectively executing a calculation process corresponding to the grip state detected by the tire grip state detection circuit 24 among the plurality of calculation processes to be performed. The vehicle body slip angle calculation circuit 30 is provided.
[0024]
In addition, the wheel speed sensor 12, the road surface μ gradient calculation circuit 14, the vibration level calculation circuit 16, and the tire grip state detection circuit 24 are disposed, for example, for each of four wheels.
[0025]
Next, the road surface μ gradient calculation circuit 14 and the vibration level calculation circuit 16 will be described in detail with reference to FIG.
[0026]
Here, the wheel speed sensor 12 outputs a sensor output signal corresponding to the wheel speed of each wheel, a so-called wheel speed sensor, and an actual rotational speed signal of each wheel (hereinafter referred to as a wheel speed signal) from the sensor output signal. And computing means for computing.
[0027]
The road surface μ gradient calculation circuit 14 detects a wheel speed vibration of each wheel as a response output of a wheel resonance system that has received a road surface disturbance from the wheel speed signal of each wheel detected by the wheel speed sensor 12. A transfer function identification circuit 34 for identifying a transfer function of each wheel that satisfies the detected wheel speed vibration using the least square method, and a μ gradient is determined for each wheel based on the identified transfer function. And an arithmetic circuit 36 for calculating.
[0028]
The pre-processing filter 32 passes only a frequency component in a certain band centered on the resonance frequency expected of the wheel resonance system, or passes only a high-frequency component including the resonance frequency component. The high-pass filter can be configured. In the present embodiment, a parameter that defines the frequency characteristics of the bandpass filter or the highpass filter is fixed to a constant value.
[0029]
Note that the output of the preprocessing filter 32 is obtained by removing the DC component. That is, only wheel speed vibration around the wheel speed signal is extracted.
[0030]
Now, the transfer function F (s) of the preprocessing filter 32 is
[0031]
[Expression 1]

[0032]
And Here, ci is a coefficient of the filter transfer function, and s is a Laplace operator.
[0033]
Next, an arithmetic expression on which the transfer function identification circuit 34 depends is derived. In the present embodiment, the calculation of the preprocessing filter 32 is included in the calculation of the transfer function identification circuit 34.
[0034]
First, the transfer function to be identified is a second-order model in which the road surface disturbance ΔTd is used as an excitation input, and the wheel speed vibration Δω1 detected by the preprocessing filter 32 at this time is a response output. That is,
[0035]
[Expression 2]

[0036]
The vibration model is assumed. Here, v is an observation noise included when the wheel speed signal is observed. When the equation (2) is transformed, the following equation is obtained.
[0037]
[Equation 3]

[0038]
First, the equation obtained by applying the preprocessing filter of equation (1) to equation (3) is discretized. At this time, Δω 1, ΔTd, v are discretized data Δω 1 (k), ΔTd (k), v (k) (k is a sampling number: k = 1, 2, 3,. ...). Further, the Laplace operator s can be discretized using a predetermined discretization method as shown in Equation 4. In this embodiment, as an example, the discretization is performed by the following bilinear transformation. Note that d is a one-sample delay operator.
[0039]
[Expression 4]

[0040]
In addition, since the order m of the preprocessing filter is preferably 2 or more, in the present embodiment, m = 2 is set in consideration of the calculation time, thereby obtaining the following equation.
[0041]
[Equation 5]

[0042]
Further, in order to identify the transfer function from each data of the wheel speed vibration Δω1 based on the least square method, the equation (4) is expressed as the following equation so as to be in the form of a linear function with respect to the parameter to be identified. Deform. “T” is the transpose of the matrix.
[0043]
[Formula 6]

[0044]
It is. In the above equation, θ is a parameter of the transfer function to be identified.
[0045]
Next, the vibration level calculation circuit 16 has a wheel speed in a predetermined range in which a frequency including at least one resonance point or at least one anti-resonance point in the wheel speed signal detected by the wheel speed sensor 12 is larger than a low frequency region. A band pass filter 38 for extracting a signal is provided. The band-pass filter 38 is connected to an arithmetic circuit 40 that calculates a physical quantity representing the magnitude of wheel vibration based on the wheel speed signal extracted from the band-pass filter 38, in this embodiment, the vibration level. Yes.
[0046]
Next, the tire grip state detection circuit 24 will be described. As shown in FIG. 3, the tire grip state detection circuit 24 includes a reference state detection circuit 42 that detects a reference state in which the grip state on the road surface of the wheel is high, and a reference state detection circuit 42 that detects the reference state. The road surface μ gradient initial value storage circuit 44 that stores the μ gradient initial value that is the average value of the μ gradient, and the μ gradient initial value stored in the road surface μ gradient initial value storage circuit 44 are compared with the detected μ gradient value. And a road surface μ gradient value comparison circuit 46 that outputs a value representing the tire grip state.
[0047]
As shown in FIG. 4, the reference state detection circuit 42 includes the wheel speeds of the left and right wheels detected by the wheel speed sensor 12, the vehicle speed detected by the vehicle speed sensor 22, the vibration level calculated by the vibration level calculation circuit 16, and the wiper. The wiper operation state detected by the operation detection sensor 20 and the environmental temperature detected by the temperature sensor 18 are input, and from these, the vehicle travels straight and at a constant speed and travels on a high μ road with a low vibration level. An AND circuit 52 that detects and outputs a signal, and when the signal is input from the AND circuit 52, a switching circuit 54 that switches the output to the μ gradient from the road surface μ gradient arithmetic circuit 14 and an output from the switching circuit 54 A high μ road detection circuit 56 that outputs a high μ road flag indicating that the wheel is in a reference state in which the grip state with respect to the road surface is high. I have.
[0048]
In the high μ road detection circuit 56, the output from the switching circuit 54 (μ gradient from the road surface μ gradient calculation circuit 14) is set to a threshold value α (maximum value of μ gradient when traveling on a high μ road determined in advance by experiments or the like). When the comparison circuit 58 determines whether or not the μ gradient from the road surface μ gradient calculation circuit 14 is larger than the threshold value α, the determination circuit 58 determines whether or not it is greater than a close value (see FIG. 8). A counter 60 that counts the determined number of times (a variable i is incremented by 1), a clock circuit 62 that resets the count value of the counter 60 when the count value is not updated for a certain period of time (i = 0), and a count of the counter 60 A comparator 64 that determines whether or not the value (i) is greater than a predetermined value N, and when the comparator 64 determines that the count value (i) of the counter 60 is greater than the predetermined value N, the value is stored in the memory 66. The A high-μ road flag, and a switcher 70 that switches to output a non-high μ road flag stored in the memory 68 when the count value (i) is not determined to be greater than the predetermined value N; It has. When the count value (i) of the counter 60 is determined to be larger than the predetermined value N, the high μ road flag stored in the memory 66 is output even if it is not a high μ road. This is because the μ gradient may be larger than the threshold value due to noise, road surface conditions, and the like, thereby preventing the high μ road flag from being output.
[0049]
As shown in FIG. 5, the road surface μ gradient initial value storage circuit 44 receives a high μ road flag and a signal from the vehicle speed sensor 22 (running start), that is, the vehicle starts traveling and the road surface of the wheel. An AND circuit 74 that outputs a signal when the grip state is high, a μ gradient average value calculating circuit 72 that averages a μ gradient within a predetermined time from when the signal from the AND circuit 74 is input, and μ A μ gradient initial value storage circuit 76 for storing the μ gradient average value calculated by the gradient average value calculation circuit 72 as the μ gradient initial value μI.
[0050]
As shown in FIG. 6, the road surface μ gradient value comparison circuit 46 includes a μ gradient average value calculation circuit 78 that calculates the average value μr of the μ gradient calculated by the road surface μ gradient calculation circuit 14 by moving average processing, and the like. A switch 80 that is turned on each time a timer signal from the timer is input, and a ratio (μr / μI) between the initial value μI of μ gradient and the average value μr of μ gradient is calculated as a value representing the tire grip state. And a μ gradient value comparison circuit 82 for output.
[0051]
Next, the operation of the present embodiment will be described. The vehicle body slip angle calculation circuit 30 according to the present embodiment has a vehicle speed determined in advance corresponding to each of a plurality of grip states and detected by the vehicle speed sensor 22, a yaw rate detected by the yaw rate sensor 26, and a steering angle. Of the plurality of calculation processes for calculating the vehicle body slip angle based on the steering angle detected by the sensor 28, the calculation process corresponding to the grip state detected by the tire grip state detection circuit 24 is selectively executed to execute the vehicle body. The slip angle is calculated.
[0052]
First, the contents detected by the tire grip state detection circuit 24 will be described. For this purpose, the calculation contents of the road surface μ gradient calculation circuit 14 and the calculation contents of the vibration level calculation circuit 16 will be described first.
[0053]
In the transfer function identification circuit 34, the unknown parameter θ is estimated by applying the least square method to each data obtained by sequentially applying the discretized data of the detected wheel speed vibration Δω1 to the equation (11). Identify the transfer function.
[0054]
Specifically, the detected wheel speed vibration Δω1 is converted into discretized data Δω (k) (k = 1, 2, 3,...), The data is sampled at N points, and the minimum The parameter θ of the transfer function is estimated using a multiplication formula.
[0055]
[Expression 7]

[0056]
Here, the amount with the symbol “^” is defined as the estimated value.
[0057]
The least square method may be calculated as a sequential least square method for obtaining the parameter θ by the following recurrence formula.
[0058]
[Equation 8]

[0059]
Here, ρ is a so-called forgetting factor, and is usually set to a value of 0.95 to 0.99. At this time, the initial value is
[0060]
[Equation 9]

[0061]
And it is sufficient.
[0062]
Various modified least square methods may be used as a method for reducing the estimation error of the least square method. In the present embodiment, an example using an auxiliary variable method that is a least square method with an auxiliary variable introduced will be described. According to this method, the parameter of the transfer function is estimated using the following equation using m (k) as an auxiliary variable when the relationship of equation (9) is obtained.
[0063]
[Expression 10]

[0064]
The sequential calculation is as follows.
[0065]
## EQU11 ##

[0066]
The principle of the auxiliary variable method is as follows. Substituting equation (15) into equation (16),
[0067]
[Expression 12]

[0068]
Therefore, if an auxiliary variable is selected so that the second term on the right side of equation (19) is zero, the estimated value of θ matches the true value of θ. Therefore, in the present embodiment, as an auxiliary variable, a variable obtained by delaying ζ (k) = [− ξy1 (k) −ξy2 (k)] T so as not to correlate with the equation error r (k) is used. . That is,
m (k) = [-ξy1 (kL) -ξy2 (kL)] T (20)
And However, L is a delay time.
[0069]
After identifying the transfer function as described above, in the arithmetic circuit 36, the physical quantity related to the road surface μ gradient D0 is
[0070]
[Formula 13]

[0071]
And calculate. Thus, when the physical quantity related to the road surface μ gradient D0 can be calculated by the equation (21), for example, when the physical quantity is small, it can be easily determined that the friction characteristic between the tire and the road surface is saturated.
[0072]
By the way, when the wheel speed signal detected by the wheel speed sensor 12 is decomposed according to the frequency, as shown in FIG. 9, it has two resonance points and one anti-resonance point. Of the two resonance points, the resonance point on the lower frequency side is a front-rear resonance point based on tire inertia or the like, and the frequency is f1 (15 to 20) Hz. The resonance point on the higher frequency side of the two resonance points is a torsional resonance point based on tire air pressure and tire rubber elasticity, and the frequency is f3 (35 to 40) Hz. The wheel speed signal has a dead zone anti-resonance point with respect to various signals, and has a frequency f2 (20 to 25) Hz. The band-pass filter 38 according to the present embodiment extracts a wheel speed signal in a predetermined range Δf including a torsional resonance point (frequency f3) from the wheel speed signals detected by the wheel speed sensor 12. Note that the bandpass filter 38 may extract a wheel speed signal in a predetermined range Δf including front and rear resonance points and antiresonance points other than the torsion resonance point.
[0073]
The arithmetic circuit 40 calculates the vibration level G (N) of the wheel speed signal determined by the following equation. The output of the band pass filter 38 is ω (k).
[0074]
[Expression 14]

[0075]
Note that the arithmetic circuit 40 actually has an arithmetic timing at each timing.
[0076]
[Expression 15]

[0077]
Are sequentially calculated. Thus, the reason why the vibration level is calculated by the arithmetic circuit 40 is that the road surface μ gradient is not properly estimated due to overpasses or the like.
[0078]
On the other hand, when the vibration level and the estimated value of the road surface μ gradient obtained in this manner are plotted on the asphalt road surface and the low μ road surface, asphalt road surface and low μ road surface are clearly distinguished as shown in FIG. In addition, even if a protrusion is passed over the asphalt road surface, it is clearly distinguished from the region on the low μ road. In view of the above facts, the present inventor obtained the relationship between the vibration level and the estimated value of the road surface μ gradient for each of various road surface conditions. As shown in FIG. As a result of experiments, it was found that each road surface condition, for example, a low μ road, an asphalt road, a cobblestone road, a snow-capped road, a gravel / non-level road, can be clearly distinguished and recognized for each region. That is, for example, the snowy road is slightly smaller than the high μ road (asphalt road, cobblestone road) as the estimated value of the road surface μ gradient, but a difference appears in the region due to the difference in vibration level. Gravel roads and the like are in a region where the vibration level is higher, but the estimated value of the road surface μ gradient is lower than that of the high μ road.
[0079]
Here, as described above, the AND circuit 52 of the reference state detection circuit 42 calculates the wheel speeds of the left and right wheels detected by the wheel speed sensor 12, the vehicle speed detected by the vehicle speed sensor 22, and the vibration level calculation circuit 16. The vibration level, the wiper operation state detected by the wiper operation detection sensor 20, and the environmental temperature detected by the temperature sensor 18 are input.
[0080]
That is, it can be determined whether the vehicle is traveling straight from the wheel speed difference between the left and right wheels detected by the wheel speed sensor 12. It can be determined whether the vehicle is traveling at a constant speed from the vehicle speed detected by the vehicle speed sensor 22. From the wiper operation state detected by the wiper operation detection sensor 20, it can be determined that it is not raining unless the wiper is operating, that is, it is a DRY road surface. Then, it can be determined whether the road surface is frozen based on the environmental temperature detected by the temperature sensor 18. Then, if the vibration level is small (for example, 0.01) and the road surface is not frozen from the vibration level calculated by the vibration level calculation circuit 16 and the environmental temperature detected by the temperature sensor 18, it is shown in FIG. From the relationship, it can be determined that the vehicle is traveling on an asphalt road surface (high μ road). When the AND circuit 52 detects that the vehicle is traveling straight and at a constant speed and traveling on a high μ road having a low vibration level, the AND circuit 52 outputs a signal.
[0081]
When the signal is input from the AND circuit 52, the switching circuit 54 switches the output to the μ gradient from the road surface μ gradient calculation circuit 14. The comparator 58 determines whether or not the output from the switching circuit 54 (the μ gradient from the road surface μ gradient calculation circuit 14) is larger than the threshold value α. When the comparison circuit 58 determines that the μ gradient from the road surface μ gradient calculation circuit 14 is larger than the threshold value α, the counter 60 counts the determined number (the variable i is incremented by 1). When the count value is not updated for a certain time or more, the count value of the counter 60 is reset (i = 0) by a signal from the clock circuit 62. The comparator 64 determines whether or not the count value (i) of the counter 60 is greater than a predetermined value N. Then, when the comparator 64 determines that the count value (i) of the counter 60 is larger than the predetermined value N, the switch 70 outputs the high μ road flag stored in the memory 66. When the count value (i) of the counter 60 is determined to be larger than the predetermined value N, the high μ road flag stored in the memory 66 is output, so that noise and road surfaces are not necessary even if the road is not a high μ road. Even if the μ gradient becomes larger than the threshold value depending on the state or the like, the high μ road flag can be prevented from being output.
[0082]
The AND circuit 74 of the road surface μ gradient initial value storage circuit 44 receives a high μ road flag and a signal from the vehicle speed sensor 22 (running start), that is, the vehicle starts to travel, and the grip state of the wheels on the road surface A signal is output when is high. The μ gradient average value calculation circuit 72 averages the μ gradient within a predetermined time from when the signal from the AND circuit 74 is input. The μ gradient initial value storage circuit 76 stores the μ gradient average value calculated by the μ gradient average value calculation circuit 72 as the μ gradient initial value μI.
[0083]
On the other hand, the μ gradient average value calculation circuit 78 of the road surface μ gradient value comparison circuit 46 calculates the average value μr of the μ gradient calculated by the road surface μ gradient calculation circuit 14 by moving average processing or the like. When the timer signal is input, the switch 80 is turned on, whereby the μ gradient average value μr calculated by the μ gradient average value calculation circuit 78 is input to the μ gradient value comparison circuit 82 and the μ gradient is input. A ratio (μr / μI) between the initial value μI and the average value μr of the μ gradient is calculated and output as a value representing the tire grip state.
[0084]
Next, a specific operation of the vehicle body slip angle calculation circuit 30 according to the present embodiment will be described with reference to the flowchart shown in FIG.
[0085]
In step 102, based on the output from the μ gradient value comparison circuit 82 arranged for each wheel, it is determined whether or not the grip level of the four wheels is equal to or higher than a first threshold value. The first threshold value is a value for determining whether or not the vehicle body slip angle is a linear state with respect to the lateral force of the wheel, and is determined in advance. That is, when the grip level of the four wheels is equal to or greater than the first threshold value, it can be determined that the vehicle body slip angle is linear with respect to the lateral force of the wheels.
[0086]
When it is determined that the grip level of the four wheels is not equal to or higher than the first threshold value, it is determined in step 104 whether or not the grip level of the four wheels is equal to or lower than the second threshold value. The second threshold value is a value for determining whether or not the vehicle is in a turning limit state where turning of the vehicle is a limit, and is determined in advance. That is, when the grip level of the four wheels is equal to or lower than the second threshold value, it can be determined that the vehicle is in a turning limit state. If a negative determination is made in step 104, the vehicle body slip angle is not a linear state with respect to the lateral force of the wheel and is an intermediate state that is not a turning limit state. Step 104 corresponds to the turning limit state detecting means in claim 4.
[0087]
When the vehicle body slip angle is linear with respect to the lateral force of the wheel, in step 106, a first slip angle estimation process (calculation process of the present invention, details will be described later) is performed. On the other hand, in the intermediate state where the vehicle body slip angle is not linear with respect to the lateral force of the wheel and is not the turning limit state, the second slip angle estimation process (calculation of the present invention) is performed in step 108. Processing and details will be described later). When the vehicle is in a turning limit state where the turning of the vehicle is a limit, in steps 110 and 112, pseudo-integration (calculation processing of the present invention, details will be described later) is performed.
[0088]
Next, the first slip angle estimation process will be described. The detected value of the yaw rate of the vehicle is γ, the feedback gain of the deviation (γ−γh) of the detected value γh of the yaw rate with respect to the estimated value βh of the body slip and the estimated value γh is Kb, and the yaw rate with respect to the estimated yaw rate γh The feedback gain of the deviation (γ−γh) between the detected value γ and the estimated value γh is Kg, the vehicle weight is m, the vehicle speed is V, and the lateral forces of the left and right front wheels and left and right rear wheels are Ff and Fr, respectively. If the distances in the longitudinal direction of the vehicle from the center of gravity of the vehicle to the front and rear wheels are Lf and Lr, respectively, the differential value βhd of the estimated value βh of the vehicle slip angle and the differential value γhd of the estimated yaw rate γh of the vehicle are as follows: This is expressed by the basic equations of Equations 16 and 17.
[Expression 16]

[0089]
[Expression 17]

If the slip angles of the front and rear wheels are βf and βr, respectively, and the cornering powers of the front and rear wheels are Cf and Cr, respectively, according to the linear theory of the vehicle model, the lateral forces Ff and Fr of the front and rear wheels are Each is expressed by the following equations 18 and 19.
[Formula 18]

[0090]
[Equation 19]

[0091]
As is well known, the cornering powers Cf and Cr are constant only when the front and rear slip angles βf and βr are relatively small, and gradually decrease as the slip angles βf and βr increase. The estimated values βfh and βrh of the front and rear slip angles βfh and βrh are the front wheel steering angle δf (which may be the average of δfr and δfl) and the rear wheel steering angle δr (the average value of δrr and δrl). It can be estimated on the basis of the estimated value βh of the slip angle of the vehicle body, distances Lf and Lr in the longitudinal direction from the center of gravity of the vehicle to the front and rear wheels, the estimated value γh of the yaw rate, and the vehicle speed V. It is represented by the formulas 20 and 21.
[0092]
[Expression 20]

[0093]
[Expression 21]

[0094]
Substituting the estimated values βfh and βrh of the equations 20 and 21 into βf and βr of the equations 18 and 19, respectively, substituting the equations 18 and 19 into the basic equations of the equations 16 and 17, and substituting the basic equations into a matrix Expressed in the form, the following formula 22 is obtained.
[0095]
[Expression 22]

[0096]
Equation 22 can be expressed as Equation 23 below.
[0097]
[Expression 23]

[0098]
Equation 23 is rewritten as Equation 24 below.
[0099]
[Expression 24]

Therefore (A-LC) ^-1 (-B * δ-γ * K) to X _C Then, the equation 24 can be rewritten as the following equation 25.
[0100]
[Expression 25]

[0101]
Equation 25 can be solved so that X converges to Xc. Since this system is a quadratic system, it has two poles p1 and p2. Therefore, βc and γc are expressed by the following equations 26 and 27 using these poles p1 and p2, respectively.
[0102]
[Equation 26]

[0103]
[Expression 27]

[0104]
In the above equations 26 and 27, βo and γo are defined as the following equation 28.
[0105]
[Expression 28]

[0106]
The feedback gains Kb and Kg are expressed by the following equations 29 and 30 using the two poles p1 and p2, respectively. Note that a11, a12, a21, and a22 are the components of the matrix A, as shown in Equation 22 above.
[0107]
[Expression 29]

[0108]
[30]

[0109]
When the wheel slip angles βf and βr are relatively small, the feedback gains Kb and Kg are used as the first feedback gain components Kb1 and Kg1, respectively, in the specification of Japanese Patent Application No. 7-10882 filed by the present applicant. In this case, the poles p1 and p2 are represented by the following equations 31 and 32, respectively, with C11, C12, C21, and C22 being positive constants.
[0110]
[31]

[0111]
[Expression 32]

[0112]
Next, the second slip angle estimation process will be described. If the slip state of the wheel advances as the vehicle spins, both the slip angle and yaw rate of the vehicle body increase, but the slip angle increases with a sign opposite to the yaw rate. Therefore, if the proportional coefficient of the yaw rate γ is set to be negative in the relationship of Equation 11, in other words, if the estimated value of the slip angle is set to increase opposite to the increasing direction of the estimated value of the yaw rate, the actual vehicle Even when the difference between the above situation and the vehicle model becomes large, the estimation error is suppressed to be small, thereby improving the accuracy of estimating the slip angle of the vehicle. Further, since the estimated value of the yaw rate must converge to the detected value of the yaw rate, the proportional coefficient of the yaw rate γ in Equation 27 must be positive. Therefore, in the non-linear region, p1 and p2 are set to satisfy the conditions of the following equations 33 and 34.
[0113]
[Expression 33]

[0114]
[Expression 34]

[0115]
The values of the poles p1 and p2 satisfying the above two conditions are shown as values in the range indicated by hatching in FIGS. In particular, FIG. 12 shows the range of poles p1 and p2 when the vehicle is traveling at low speed, and FIG. 13 shows the range of poles p1 and p2 when the vehicle is traveling at high speed. The ranges of the poles p1 and p2 gradually change from the range of FIG. 12 to the range of FIG. 13 as the vehicle speed increases.
[0116]
According to the nonlinear theory of the vehicle model, the lateral forces Ff and Fr of the front and rear wheels are represented by the graph shown in FIG. The slip angles βf and βr of the front and rear wheels increase to the vicinity of the value indicated by, for example, βm in FIG. 14, and the front and rear slip angles βf and βr and the lateral forces Ff and If the proportional relationship with Fr is greatly lost, the estimation of the slip angle of the car body is further corrected.
[0117]
Although the cornering powers Cf and Cr change from moment to moment, the lateral forces Ff and Fr of the front and rear wheels can be expressed by the following equations 20 and 21 for convenience. In the following equations 35 and 36, ΔFf and ΔFr are estimation errors including an error in cornering power Cp and an estimation error in the friction coefficient μ of the road surface. Ff0 and Fr0 are not constants but functions of βfh and βrh, respectively. However, if only the vicinity of the current value βm of βfh and βrh is considered, these can be approximated to constants.
[0118]
[Expression 35]

[0119]
[Expression 36]

[0120]
Substituting Ff and Fr in Equations 35 and 36 into the above Equations 31 and 32, and expressing the results in matrix form, the following Equation 37 is obtained.
[0121]
[Expression 37]

[0122]
When the equation 37 is Laplace transformed, the following equation 38 is obtained.
[0123]
[Formula 38]

[0124]
Since the lateral force estimated in the non-linear region of the tire lateral force (however, in the above intermediate state) includes a large error, when the tire grip of the front wheels is saturated by adjusting the feedback coefficient α The transfer function from ΔFf to βh is set to 0, and when the rear wheel tire grip is in a saturated state, the transfer function from ΔFr to βh is set to 0, thereby estimating the slip angle β of the vehicle due to lateral force estimation error. Errors are reduced.
[0125]
The feedback coefficient α for setting the transfer function from ΔFf to βh to 0 is expressed by the following equation (39), and the feedback coefficient α for setting the transfer function from ΔFr to βh to 0 is expressed by the following equation (40). Is done.
[0126]
[39]

[0127]
[Formula 40]

[0128]
Next, the pseudo integration will be described. The pseudo-integration is an error reduction calculation process for calculating a vehicle body slip angle that reduces the influence of the error due to the turning limit state.
By ignoring the physical quantity used to calculate the vehicle body slip angle last time and integrating the physical quantity newly detected to calculate the vehicle body slip angle this time, an error caused by the turning limit state. To reduce the effects of
[0129]
That is, in step 110, dβ = Gy / Vx−γ is calculated from the physical quantity of the longitudinal speed Vx approximated from the detected vehicle speed, the lateral acceleration Gy acting on the vehicle, and the detected yaw rate γ, and step 112 Then, βest = T / (1 + T * s) dβ is pseudo-integrated. That is, ignoring the physical quantity used to calculate the vehicle body slip angle last time, for example, 20 [μS] elapses after this step 112 is newly started to calculate the vehicle body slip angle this time. The detected physical quantity is integrated. Note that s is discretized by using the Laplace operator using Equation (4) for mounting on a computer.
[0130]
After the first slip angle estimation process, if the slip angle β1 obtained by the first slip angle estimation process is 0 or more in step 114, the slip angle β1 and the predetermined value A0 are set as the slip angle β1. When the slip angle β1 obtained by the first slip angle estimation process is smaller than 0, the slip angle β1 is determined as a slip angle β1 between the slip angle β1 and a negative value of the predetermined value A0. Take a large value.
[0131]
As described above, in the present embodiment, a plurality of (three) calculation processes for calculating the vehicle body slip angle are determined in advance for each of a plurality of grip states with respect to the road surface of the wheel, and the detected grip state is set. Since the corresponding calculation process is selectively executed, the vehicle body slip angle can be accurately calculated regardless of the grip state of the wheels with respect to the road surface.
[0132]
In addition, when a turning limit state is detected, an error reduction calculation process for calculating a vehicle body slip angle that reduces the influence of the error caused by the turning limit state is executed. Even in the turning limit state, the vehicle body slip angle can be accurately calculated. In this case, the physical quantity used for the previous calculation of the vehicle body slip angle is ignored, and the newly detected physical quantity is integrated to calculate the vehicle body slip angle this time. Since the vehicle body slip angle with reduced influence of the error is calculated, it is possible to eliminate the influence of the past error included in the physical quantity and due to the turning limit state. In addition, the vehicle body slip angle can be accurately calculated.
[0133]
When the linear state in which the vehicle body slip angle is linear with respect to the lateral force of the wheel is detected and the calculated vehicle body slip angle exceeds the limited value, the calculated vehicle body slip angle is restricted to the limited value. Thus, deterioration of detection accuracy can be prevented.
[0134]
In the embodiment described above, the tire grip state (turning limit state) is calculated using the road surface μ gradient, but the present invention is not limited to this, and the slip speed equivalent to the road surface μ gradient is calculated. The same calculation may be performed using a braking torque gradient that is a braking torque gradient and a driving torque gradient that is a driving torque gradient. That is, the tire grip state can be calculated based on physical quantities representing the ease of slipping of the wheel, such as road surface μ gradient, braking torque gradient, and drive torque gradient.
[0135]
【The invention's effect】
As described above, the present invention predetermines a plurality of calculation processes for calculating the vehicle body slip angle corresponding to each of a plurality of grip states with respect to the road surface of the wheel, and selects a calculation process corresponding to the detected grip state. Therefore, the vehicle body slip angle can be accurately calculated regardless of the grip state of the wheels with respect to the road surface.
[0136]
According to a second aspect of the present invention, when a turning limit state is detected, a vehicle body slip angle that is included in the physical quantity detected by the physical quantity detection means and that reduces the influence of errors caused by the turning limit state. Therefore, the vehicle body slip angle can be accurately calculated even in a turning limit state in which the influence of the error is sometimes great.
[0137]
The invention according to claim 3 ignores the physical quantity used to calculate the vehicle body slip angle last time, and integrates the newly detected physical quantity to calculate the vehicle body slip angle this time. Since the vehicle slip angle is calculated by reducing the influence of the error due to certain things, it is possible to eliminate the influence of past errors included in the physical quantity and due to the turning limit state, and the turning limit state where the influence of the error is sometimes large In this case, the vehicle body slip angle can be accurately calculated.
[0138]
In the invention according to claim 4, when a linear state in which the vehicle body slip angle is linear with respect to the lateral force of the wheel is detected and the calculated vehicle body slip angle exceeds a limit value, the calculated vehicle body slip angle is Since it is restricted to the limit value, there is an effect that deterioration of detection accuracy can be prevented.
[0139]
According to the fifth aspect of the present invention, when a turning limit state is detected, the vehicle body slip angle is included in the physical quantity detected by the physical quantity detection means and the influence of an error caused by the turning limit state is reduced. Therefore, the vehicle body slip angle can be accurately calculated even in a turning limit state where the influence of the error is sometimes great.
[Brief description of the drawings]
FIG. 1 is a block diagram of a vehicle body slip calculation circuit according to a first embodiment of the present invention.
FIG. 2 is a block diagram of a road surface μ gradient calculation circuit and a vibration level calculation circuit.
FIG. 3 is a block diagram of a tire grip state detection circuit.
FIG. 4 is a diagram illustrating a reference state detection circuit.
FIG. 5 is a diagram showing a road surface μ gradient initial value storage circuit;
FIG. 6 is a diagram showing a road surface μ gradient value comparison circuit;
FIG. 7 is a flowchart showing the operation of the present embodiment.
FIG. 8 is an explanatory diagram for explaining a threshold value in the reference state detection circuit;
FIG. 9 is a graph showing the relationship between the frequency and amplitude of a wheel speed signal.
FIG. 10 is a graph showing the relationship between the vibration level of the wheel speed signal and the road surface μ gradient for asphalt roads and low μ roads.
FIG. 11 is a graph showing the relationship between the vibration level of the wheel speed signal and the road surface μ gradient for each road surface state.
FIG. 12 is a graph showing an appropriate range of second feedback gain calculation parameters p1 and p2 when the vehicle is traveling at a low speed.
FIG. 13 is a graph showing appropriate ranges of second feedback gain calculation parameters b1 and p2 when the vehicle is traveling at high speed.
FIG. 14 is a graph showing a relationship between slip angles βf and βr and lateral forces Ff and Fr of front wheels and rear wheels.
[Explanation of symbols]
24 Tire grip state detection circuit (grip state detection means, turning limit state detection means)
22 Vehicle speed sensor (physical quantity detection means)
26 Yaw rate sensor (physical quantity detection means)
28 Rudder angle sensor (physical quantity detection means)
30 Vehicle slip angle calculation circuit (calculation execution means, vehicle slip angle calculation means)

Claims (7)

Grip state detection means for detecting the grip state of the wheel with respect to the road surface using the road surface μ gradient ,
Physical quantity detection means for detecting a physical quantity representing the behavior of the vehicle body necessary for calculating the vehicle body slip angle;
The slip angle of the vehicle body is estimated based on the physical quantity detected in advance by the physical quantity detecting means corresponding to each grip state and using the lateral force of the front and rear wheels in the equation of motion of the vehicle as a linear characteristic. A plurality of calculation processes for calculating the vehicle body slip angle, including a first slip angle estimation process and a second slip angle estimation process for estimating the slip angle of the vehicle body using the lateral force of the front and rear wheels in the vehicle equation of motion as nonlinear characteristics . Calculation execution means for selectively executing calculation processing corresponding to the grip state detected by the grip state detection means,
A vehicle body slip angle calculation device. The grip state includes a turning limit state where turning of the vehicle is a limit,
The calculation execution means further includes an error reduction calculation process for calculating a vehicle body slip angle that is included in the physical quantity detected by the physical quantity detection means and that reduces an influence of an error caused by the turning limit state. , when the turning limit state is detected by the grip state detection means executes the error reduction processing,
The vehicle body slip angle calculating device according to claim 1. The calculation execution means, when executing the error reduction calculation process,
By ignoring the physical quantity used to calculate the vehicle body slip angle last time and integrating the newly detected physical quantity to calculate the vehicle body slip angle this time, it results from the turning limit state. 3. The vehicle body slip angle calculation device according to claim 2, wherein the vehicle body slip angle with reduced influence of errors is calculated. The grip state includes a linear state in which the vehicle body slip angle is linear with respect to the lateral force of the wheel,
Said execution means, if the vehicle body slip angle is calculated slip angle and computation by said first slip angle estimation process in the case where the linear state is detected by the grip condition detecting means exceeds a limit value Is characterized by restricting the calculated vehicle body slip angle to a limited value,
The vehicle body slip angle calculating device according to any one of claims 1 to 3. A turning limit state detecting means for detecting a turning limit state where turning of the vehicle is a limit using a road surface μ gradient ;
Physical quantity detection means for detecting a physical quantity representing the behavior of the vehicle body necessary for calculating the vehicle body slip angle;
When the turning limit state is detected by the turning limit state detection means, based on the physical quantity detected by the physical quantity detection means, an error caused by being in the turning limit state and being included in the physical quantity A vehicle body slip angle calculating means for calculating a vehicle body slip angle with reduced influence;
A vehicle body slip angle calculation device. The grip state detection means calculates a ratio between the initial value of the road surface μ gradient and the average value of the μ gradient obtained by calculating the average value of the μ gradient from the time when the vehicle starts running and the grip state is detected in a predetermined high state The vehicle body slip angle according to any one of claims 1 to 4, wherein the grip state is detected by calculating as a value representing the grip state and comparing with a predetermined threshold value for each grip state. Arithmetic unit. The turning limit state detecting means is a ratio between the initial value of the road surface μ gradient and the average value of the μ gradient obtained by calculating the average value of the μ gradient from the time when the vehicle starts traveling and the grip state is detected in a predetermined high state. 6. The vehicle body slip angle calculation device according to claim 5, wherein the turning limit state is detected by calculating and comparing with a predetermined threshold value for detecting the turning limit state.

Images