JP3617309B2 - Road friction coefficient estimation device - Google Patents

Description

ただし、ｘ_ｗ は車輪速、ｘ_ｖ は車体速
で表わされる、路面およびタイヤ間のスリップ率λの関数であり、該スリップ率λが所定の負の定数−λ_１ 以上かつ所定の正の数λ_２ 未満の場合に単調に増加し、該スリップ率λが所定の負の定数−λ_１ 未満または所定の正の数λ_２ 以上の場合に単調に減少するものであり、
前記疑似μ−ｓ関数をμ_ｑｕとすると、μ_ｑｕは、次式
μ_ｑｕ＝Ｄsin ［Ｃtan ^−１（Ｂφ）］ (2)
φ＝１００（１−Ｅ）λ＋（Ｅ／Ｂ）tan （１００Ｂλ） (3)
ただし、Ｂ，Ｃ，Ｄ，Ｅはタイヤ特性によって決定される定数
で表わされるマジックフォーミュラにより演算することを特徴とするものである。
【０００８】
本発明の請求項２に係る路面摩擦係数推定装置は、車輪速を検出する車輪速検出手段と、車体速を検出する車体速検出手段と、車輪の制動力および駆動力を検出する制駆動力検出手段と、前記車輪速および車体速を入力されて車輪のスリップ率を演算するスリップ率演算手段と、前記スリップ率を入力されて路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数を発生する疑似μ−ｓ関数発生手段と、前記車輪速および制駆動力ならびに路面摩擦係数推定値を入力されて該路面摩擦係数推定値に応じた車輪速推定値を演算するとともに、該車輪速推定値を前記車輪速から減じて車輪速推定誤差を演算する車輪速推定誤差演算手段と、前記車輪速推定誤差および疑似μ−ｓ関数を入力されて該疑似μ−ｓ関数を校正するための校正パラメータを演算する校正パラメータ演算手段と、前記疑似μ−ｓ関数および校正パラメータを入力されて該疑似μ−ｓ関数を前記校正パラメータで校正し、その校正値を前記路面摩擦係数推定値として出力する校正手段とから成り、前記車輪速推定誤差演算手段は、車輪速推定誤差の単調非減少関数と制駆動力と路面摩擦係数推定値との線形結合
（ここで、路面摩擦係数推定値の結合係数は輪荷重を車輪慣性モーメントで割った値、制駆動力の結合定数は車輪慣性モーメントの逆数、車輪速推定誤差もしくは車輪速推定誤差の正負符号の結合定数は適宜決定する所定の負の値）
を当該車輪速推定値の変化量とすることを特徴とするものである。
【０００９】
本発明の請求項３に係る路面摩擦係数推定装置は、上記請求項２記載の路面摩擦係数推定装置における車輪速推定誤差の単調非減少関数は、車輪速推定誤差と適宜決定する所定のゲインとの積もしくは車輪速推定誤差の正負符号と適宜決定する所定のゲインとの積であることを特徴とするものである。
【００１０】
本発明の請求項４に係る路面摩擦係数推定装置は、車輪速を検出する車輪速検出手段と、車体速を検出する車体速検出手段と、車輪の制動力および駆動力を検出する制駆動力検出手段と、前記車輪速および車体速を入力されて車輪のスリップ率を演算するスリップ率演算手段と、前記スリップ率を入力されて路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数を発生する疑似μ−ｓ関数発生手段と、前記車輪速および制駆動力ならびに路面摩擦係数推定値を入力されて該路面摩擦係数推定値に応じた車輪速推定値を演算するとともに、該車輪速推定値を前記車輪速から減じて車輪速推定誤差を演算する車輪速推定誤差演算手段と、前記車輪速推定誤差および疑似μ−ｓ関数を入力されて該疑似μ−ｓ関数を校正するための校正パラメータを演算する校正パラメータ演算手段と、前記疑似μ−ｓ関数および校正パラメータを入力されて該疑似μ−ｓ関数を前記校正パラメータで校正し、その校正値を前記路面摩擦係数推定値として出力する校正手段とから成り、前記校正パラメータ演算手段は、校正パラメータの変化率を車輪推定誤差と疑似μ−ｓ関数値と正の定数との積により演算することを特徴とするものである。
【００１１】
【発明の効果】
本発明の請求項１によれば、車輪速および車体速から演算された車輪のスリップ率に基づき、路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数が発生され、車輪速および制駆動力ならびに路面摩擦係数推定値に基づき、該路面摩擦係数推定値に応じた車輪速推定値および車輪速推定誤差が演算され、前記車輪速推定誤差および疑似μ−ｓ関数に基づき該疑似μ−ｓ関数の校正に用いる校正パラメータが演算され、この校正パラメータで前記疑似μ−ｓ関数を校正した校正値が前記路面摩擦係数推定値として出力されるから、該路面摩擦係数推定値は速やかに路面摩擦係数の真値に収束することになる。したがって、路面摩擦係数値自体を精度良く推定し得る路面摩擦係数推定装置を提供することができる。
また、本発明の請求項１によれば、前記疑似μ−ｓ関数は、路面およびタイヤ間のスリップ率λが所定の負の定数−λ_１ 以上かつ所定の正の数λ_２ 未満の場合に単調に増加し、該スリップ率λが所定の負の定数−λ_１ 未満または所定の正の数λ_２ 以上の場合に単調に減少するから、この疑似μ−ｓ関数は路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数となる。よって、この疑似μ−ｓ関数を前記校正パラメータで校正することにより、路面摩擦係数をスリップ率により表わした関数を得ることができる。
さらに、本発明の請求項１によれば、前記疑似μ−ｓ関数μ_ｑｕは、上記（２），（３）式で表わされるマジックフォーミュラにより演算するから、この疑似μ−ｓ関数は上記路面摩擦係数の推定に適した関数となる。
【００１２】
本発明の請求項２によれば、車輪速および車体速から演算された車輪のスリップ率に基づき、路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数が発生され、車輪速および制駆動力ならびに路面摩擦係数推定値に基づき、該路面摩擦係数推定値に応じた車輪速推定値および車輪速推定誤差が演算され、前記車輪速推定誤差および疑似μ−ｓ関数に基づき該疑似μ−ｓ関数の校正に用いる校正パラメータが演算され、この校正パラメータで前記疑似μ−ｓ関数を校正した校正値が前記路面摩擦係数推定値として出力される際に、前記車輪速推定誤差の単調非減少関数と制駆動力と路面摩擦係数推定値との線形結合を当該車輪速推定値の変化量とするから、公知のように車輪速推定値の振動を抑制することができる。
【００１３】
本発明の請求項３によれば、前記車輪速推定誤差の単調非減少関数は、車輪速推定誤差と適宜決定する所定のゲインとの積もしくは車輪速推定誤差の正負符号と適宜決定する所定のゲインとの積により決定されるから、この車輪速推定誤差の単調非減少関数は上記路面摩擦係数の推定に適した関数となる。
【００１４】
本発明の請求項４によれば、車輪速および車体速から演算された車輪のスリップ率に基づき、路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数が発生され、車輪速および制駆動力ならびに路面摩擦係数推定値に基づき、該路面摩擦係数推定値に応じた車輪速推定値および車輪速推定誤差が演算され、前記車輪速推定誤差および疑似μ−ｓ関数に基づき該疑似μ−ｓ関数の校正に用いる校正パラメータが演算され、この校正パラメータで前記疑似μ−ｓ関数を校正した校正値が前記路面摩擦係数推定値として出力される際に、前記校正パラメータの変化率は車輪推定誤差と疑似μ−ｓ関数値と正の定数との積により演算されるから、この校正パラメータの変化率は当該校正パラメータの算出に適したものとなる。
【００２１】
【発明の実施の形態】
以下、本発明の実施の形態を図面に基づき詳細に説明する。
図１は本発明の路面摩擦係数推定装置の原理的構成を示す図である。本発明の路面摩擦係数推定装置は、図１に示すように、車輪速検出手段である車輪速センサ１、車体速検出手段である車体速センサ２、制駆動トルク検出手段３、スリップ率演算手段４、疑似μ−ｓ関数発生手段５、車輪速推定誤差演算手段６、校正パラメータ演算手段７および校正手段８を具備して成る。
【００２２】
車輪速センサ１は、車輪速（車輪の回転速度）ｘ_ｗ を検出する。この車輪速センサ１としては、タコジェネレータや回転型インクリメンタルエンコーダの発生パルスを周期計測する装置を用いることができる。
車体速センサ２は、車体速ｘ_ｖ を検出する。この車体センサ２としては、非駆動輪または非制動輪の回転速度を検出して、該検出値を車体速とするものを用いることができる。なお、上記の代わりに、空間フィルタを用いた光学式対地車速計やドップラレーダ等を利用してもよい。
【００２３】
制駆動力検出手段３は、車輪の制駆動力であるブレーキトルクや駆動トルク（以下、制駆動トルクという）ｕを検出する。この制駆動トルクに関しては、制動時にはブレーキ液圧を、駆動時にはエンジントルクを概略制駆動トルク相当値と見なすことができる。例えば、エンジントルク推定値ｑ_ｅ をエンジン制御装置から入力するとともにブレーキ液圧ｐ_ｂ をブレーキシリンダに取り付けた圧力センサから入力したとき、次式
ｕ＝ｃ_ｅ ｑ_ｅ ＋ｃ_ｂ ｐ_ｂ （４）
ここで、ｃ_ｅ はそれぞれエンジントルクから制駆動トルクへの換算係数、ｃ_ｂ はブレーキ液圧から制駆動トルクへの換算係数
によって制駆動トルクｕを算出する。
【００２４】
スリップ率演算手段４は、車輪速ｘ_ｗ および車体速ｘ_ｖ を入力されて、車輪のスリップ率λを、次式
【数３】

により演算する。なお、上記車体速ｘ_ｖ は、ｘ_ｖ ≒Ｖ_ｖ ／ｒ_ｗ により車体の並進速度Ｖ_ｖ および車輪半径ｒ_ｗ から求めることができる。
【００２５】
疑似μ−ｓ関数発生手段５は、スリップ率λを入力されて、路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数に相似な関数である疑似μ−ｓ関数μ_ｑｕを、次式
【数４】

により演算する。この疑似μ−ｓ関数μ_ｑｕは、例えば図３に示すように、正の傾きｋ_２ を有する１次関数の両端に負の傾き−ｋ_１ を有する１次関数を接続した形状となり、図２に例示するようなμ−ｓ関数とは形状が相似することになる。
【００２６】
疑似μ−ｓ関数発生手段５は、上記（５）式に代えて、次式で表わされるマジックフォーミュラ
μ_ｑｕ＝Ｄｓｉｎ ［Ｃｔａｎ ^−１（Ｂφ）］ （２）
φ＝１００（１−Ｅ）λ＋（Ｅ／Ｂ）ｔａｎ （１００Ｂλ） （３）
により疑似μ−ｓ関数μ_ｑｕを演算することもできる。このマジックフォーミュラ中のＢ，Ｃ，Ｄ，Ｅに適当な係数を与えると、実際のμ−ｓ関数を高い精度で近似できることが知られている（参考文献：Ｈ．Ｂ．Ｐａｃｅｊｋａ， Ｔｉｒｅ ｍｏｄｅｌｓ ｆｏｒ ｖｅｈｉｃｌｅ ｄｙｎａｍｉｃｓ ａｎａｌｙｓｉｓ， Ｓｕｐｐｌｅｍｅｎｔ ｔｏ Ｖｅｈｉｃｌｅ Ｓｙｓｔｅｍ Ｄｙｎａｍｉｃｓ， ｖｏｌ．２１，（１９９１）を参照のこと）。本実施形態では、例えばＢ＝０．１、Ｃ＝１．５５、Ｄ＝１、Ｅ＝０．４３２を与えるものとし、その場合には疑似μ−ｓ関数μ_ｑｕは図４に示す形状になる。
【００２７】
車輪速推定誤差演算手段６は、制駆動トルクｕ、車輪速ｘ_ｗ および後述する路面摩擦係数推定値μ_ｐｒを入力されて、次式
【数５】

ただし、Ｗは輪荷重、Ｉは車輪慣性モーメント、ｈ_１ およびｈ_２ は設計者が適宜決定する正の定数、ｓｇｎ（ｘ_ｗｐｒ −ｘ_ｗ ） は（ ｘ_ｗｐｒ −ｘ_ｗ ） の正負符号であり、−１，０，１の何れかの値となる
により車輪速推定値ｘ_ｗｐｒ を演算する。そして、車輪速推定値ｘ_ｗｐｒ および車輪速ｘ_ｗ から車輪速推定誤差ｅを、次式
ｅ＝ｘ_ｗ −ｘ_ｗｐｒ （７）
により演算する。
【００２８】
校正パラメータ演算手段７は、車輪速推定誤差ｅおよび疑似μ−ｓ関数μ_ｑｕを入力されて、次式
【数６】

ただし、ｃは設計者が適宜決定する正の定数
により疑似μ−ｓ関数μ_ｑｕを校正するための校正パラメータθ_ｐｒを演算する。
【００２９】
校正手段８は、疑似μ−ｓ関数μ_ｑｕおよび校正パラメータθ_ｐｒを入力されて、次式
μ_ｐｒ＝θ_ｐｒμ_ｑｕ （９）
により該疑似μ−ｓ関数μ_ｑｕを校正パラメータθ_ｐｒで校正する演算を行い、その校正値を路面摩擦係数推定値μ_ｐｒとして出力する。
【００３０】
次に、本実施形態における路面摩擦係数推定原理について説明する。
路面およびタイヤ間の摩擦係数は、タイヤおよび路面間のスリップ率に依存して変化するが、路面摩擦係数μをスリップ率λの関数として表わす関数μ（λ）は、図２に示すように路面の種類（乾燥路、湿潤路、氷結路）に拘わらず相似形状となる。そこで、未知の校正パラメータをθとしたとき、路面摩擦係数を表わす関数μ（λ）は、既知の関数である疑似μ−ｓ関数μ_ｑｕ（λ）と未知の校正パラメータθとにより、概略
μ（λ）≒θμ_ｑｕ（λ） （１０）
と近似することができる。
【００３１】
よって、μの推定値μ_ｐｒを得るためには、θの推定値θ_ｐｒを用いて、次式
μ_ｐｒ＝θ_ｐｒμ_ｑｕ （１１）
の演算を行えばよい。ここで、推定値θ_ｐｒは疑似μ−ｓ関数μ_ｑｕを校正するパラメータであるので、以下、この推定値θ_ｐｒ自体を校正パラメータと呼ぶことにする。
【００３２】
上記路面摩擦係数推定値μ_ｐｒは、次式
【数７】

と表わされるものと考える。この場合、μは有限の値を取るので、
【外１】

も有限の値となる。すなわち、
【数８】

となる正の定数γ_０ が存在する。
【００３３】
さて、車輪の運動方程式は、
【数９】

ただし、ｘ_ｗ は車輪速（車輪の回転速度）、Ｗは輪荷重、ｕは制駆動トルク、Ｉは車輪慣性モーメント
と表わすことができる。
【００３４】
この（１５）式に基づいて、以下の演算により校正パラメータθ_ｐｒを得ることができる。
【数１０】

ここで、

ただし、ｘ_ｗｐｒ はｘ_ｗ の推定値、ｈ_１ ，ｈ_２ およびｃは設計者が適宜決定する正の定数であり、ｈ_２ は特に、
ｈ_２ ＞γ_０ Ｗ／Ｉ （１７）
ただし、γ_０ は正の定数
を満たすように選定する。
【００３５】
校正パラメータθ_ｐｒが時間とともにθに収束することは、以下のように示すことができる。
（１５）式の両辺から（６）式の両辺をそれぞれ引くと、
【数１１】

を得る。
【００３６】
一方、誤差ｅおよび（θ−θ_ｐｒ） の収束性を簡単な方法で保証するために、次式で表わされる、正の値を取る関数Ｖを想定し、
Ｖ＝（１／２）ｅ^２ ＋（ｃ／２）（θ−θ_ｐｒ）^２ （１９）
その時間変化を考えると、θを定数と仮定したとき、Ｖの時間変化 ｄＶ／ｄｔは、
【数１２】

を得る。
【００３７】
ここで、（８）式によりθを変化させれば右辺第２項は０となり、ｈ_１ ＞０，ｈ_２ ＞γ_０ Ｗ／Ｉであるので、Ｖの時間変化ｄＶ／ｄｔは、次式
【数１３】

となり、関数Ｖは単調減少して０に収束する。すなわち、ｅおよび （θ−θ_ｐｒ） も０に収束するので、θ_ｐｒはθに収束する。よって、μ_ｐｒ＝θ_ｐｒμ_ｑｕはμに収束する。
【００３８】
上記（２１）式より明らかなように、ｈ_１ を０として（６）式のｈ_１ に関する項を省略することもできる。ただし、ｈ_１ およびｈ_２ を大きな値とするほど路面摩擦係数の推定の速度が向上し、ｈ_２ を大きな値とするほど
【外２】

が大きくなっても正しい路面摩擦係数を推定しやすくなる。しかし、ｈ_２ の値が過大になると、車輪速ｘ_ｗ の検出値にノイズが混在している場合にそのノイズの影響を受け、推定値が振動することがある。そこで、本実施形態ではｓｇｎ（ｘ_ｗｐｒ −ｘ_ｗ ）を以下のような単調非減少な連続関数ｆ_１ またはｆ_２ で代用するようにしており、そのようにすると車輪速推定値の振動が抑制されることが知られている。
ｆ_１（ｘ_ｗｐｒ −ｘ_ｗ ） ：＝｜ｘ_ｗｐｒ −ｘ_ｗ ｜／（ｘ_ｗｐｒ −ｘ_ｗ ＋ε） （２３）
ここで、「：＝」は「左辺を右辺と定義する」ことを意味し、εは正の定数である。
【数１４】

ここで、ｆ_０ およびη_０ は設計者が決定する正の定数である。
【００３９】
本実施形態の路面摩擦係数推定装置によれば、以下の効果が得られる。
（１）何れかの車輪に滑りが発生していれば、路面摩擦係数が推定可能である。
（２）速やかかつ正確に路面摩擦係数を推定することができる。
（３）車輪の慣性モーメントＩおよび輪荷重Ｗが既知であれば、どのような種類の車輪にも適用可能である。
（４）車輪速および車体速を検出することができれば、路面摩擦係数推定装置が構成可能である。
（５）校正パラメータθが大きいことは滑りにくい路面であることを意味し、校正パラメータθが小であることは滑り易い路面であることを意味するため、θから路面の滑り易さを定量的に判定することができる。
【００４０】
【実施例】
以下、本発明の実施例を図面に基づき詳細に説明する。
図５は本発明の第１実施例の路面摩擦係数推定装置を車両に搭載した場合の構成を示す図である。本実施例においては、タイヤ（図５には１輪のみを代表的に示してある）１１の近傍にブレーキ液圧センサ１２および車輪速センサ１３を設けてこれらセンサ１２，１３を計算機ユニット１４に接続するとともに、この計算機ユニット１４にはエンジン１５のエンジン制御装置１６および車体速センサ１７を接続する。本実施例では、上記車輪速センサ１３としてロータリエンコーダを用いるとともに、上記車体速センサ１７として対地車速計を用いており、ブレーキ液圧センサ１２の検出値およびエンジン制御装置１６から出力されるエンジントルク推定値に基づいて制駆動トルクを演算する。
【００４１】
上記計算機ユニット１４は、上記ブレーキ液圧センサ検出値およびエンジントルク推定値から制駆動トルクを演算する処理を行うとともに、上述したスリップ率演算手段、疑似μ−ｓ関数発生手段、車輪速推定誤差演算手段、校正パラメータ演算手段および校正手段に対応する処理を行う。これらの処理は、車輪速ｘ_ｗ の変化に対し十分短く設定したサンプル時間（例えば１０［ｍｓｅｃ］）毎に行われる。
【００４２】
図６は第１実施例における計算機ユニットによる路面摩擦係数推定処理を示すフローチャートである。図６において、まず、ステップ５１では、サンプル時刻になったか否かを判定し、サンプル時刻である場合に限り次のステップ５２でブレーキ液圧センサ１２からブレーキ液圧ｐ_ｂ を読み込む。次いで、ステップ５３，５４では、車輪速センサ１３から車輪速ｘ_ｗ を読み込むとともに車体速センサ１７から車体速ｘ_ｖ を読み込む。
【００４３】
次のステップ５５では、スリップ率λを、次式
【数１５】

により演算する。
次のステップ５６では、疑似μ−ｓ関数μ_ｑｕを、次式
μ_ｑｕ＝μ_ｑｕ （λ） （２５）
により演算し、次いで、ステップ５７で、路面摩擦係数推定値μ_ｐｒを、次式
μ_ｐｒ＝θ_ｐｒμ_ｑｕ （９）
により演算する。
【００４４】
次のステップ５８では、次式
【数１６】

により車輪速推定誤差ｅを演算し、次いで、ステップ５９で、校正パラメータθ_ｐｒを、次式
【数１７】

により演算する。
【００４５】
図７（ａ）〜（ｃ）は第１実施例による路面摩擦係数の推定結果を示す図である。この図７（ａ）〜（ｃ）は、自動車が加速するときの路面摩擦係数を推定する際の状況を示している。すなわち、駆動トルクｕが上昇するとそれに応じてスリップ率λも上昇し、路面摩擦係数μおよびその推定値μ_ｐｒも変化するが、本実施例では、一旦路面摩擦係数推定値μ_ｐｒが真値μに収束すると、路面摩擦係数μが変化しても精度のよい路面摩擦係数推定値μ_ｐｒが得られる。また、路面摩擦係数μの収束は、非常に速やかである。
【００４６】
本実施例は上述した例のみに限定されるものではなく、種々の変更または変形を加えることができる。例えば、自動車においては通常４つの車輪速検出値を得ることができるが、該４つの車輪速検出値の中から、駆動時には最も小さいもの、制動時には最も大きいものを選択して車体速検出値として代用すると、路面摩擦係数の推定精度は若干低下するが、装置構成を単純化することができる。
【００４７】
また、スリップ率の絶対値が小さく、したがって摩擦力自体の絶対値が小さいときは、演算精度が低下することがある。そこで、スリップ率の絶対値が大きいときのθの値に重み付けを行って、路面摩擦係数を推定してもよい。すなわち、路面摩擦係数推定値μ_ｐｒを、次式
【数１８】

ただし、Ｗ（λ）はλの増加関数
により演算するようにしてもよい。
【図面の簡単な説明】
【図１】本発明の路面摩擦係数推定装置の原理的構成を示す図である。
【図２】路面およびタイヤ間に発生する摩擦係数をスリップ率の関数として表わすμ−ｓ関数を例示する図である。
【図３】本発明の第１実施形態における疑似μ−ｓ関数の一例を示す図である。
【図４】本発明の第１実施形態における疑似μ−ｓ関数の他の例を示す図である。
【図５】本発明の第１実施例の路面摩擦係数推定装置を車両に搭載した場合の構成を示す図である。
【図６】図６は第１実施例における計算機ユニットによる路面摩擦係数推定処理を示すフローチャートである。
【図７】（ａ）〜（ｃ）は第１実施例による路面摩擦係数の推定結果を示す図である。
【符号の説明】
１ 車輪速センサ（車輪速検出手段）
２ 車体速センサ（車体速検出手段）
３ 制駆動トルク検出手段（制駆動力検出手段）
４ スリップ率演算手段
５ 疑似μ−ｓ関数発生手段
６ 車輪速推定誤差演算手段
７ 校正パラメータ演算手段
８ 校正手段[0001]
BACKGROUND OF THE INVENTION
The present invention is used in a control system for controlling the slip ratio of a vehicle wheel such as an anti-lock brake control device or a traction control device, and can accurately estimate a road surface friction coefficient estimated value. The present invention relates to an estimation apparatus.
[0002]
[Prior art]
Examples of conventional techniques for estimating the road surface friction coefficient include the following a) to f).
a) a device for estimating a road surface friction coefficient when the wheel rotational acceleration is close to 0 (see Japanese Patent Laid-Open No. 5-77706),
b) a device for estimating a road surface friction coefficient from a torque necessary for limiting the differential between the left and right wheels (see Japanese Patent Laid-Open No. 5-346394),
c) an apparatus for estimating a road surface friction coefficient from the relationship between a slip angle during steering and a cornering force (see JP-A-5-255510),
d) a device for estimating the road surface friction coefficient from the total acceleration generated in the vehicle body (see Japanese Patent Laid-Open No. 6-92218),
e) an apparatus for estimating the road surface friction coefficient from the stress of the suspension (see Japanese Patent Laid-Open No. 6-288798),
f) A method of estimating the road surface friction coefficient from the operating state of the antilock brake (see Japanese Patent Laid-Open Nos. 5-208670 and 9-263226).
[0003]
[Problems to be solved by the invention]
In the apparatus a), since the road surface friction coefficient cannot be estimated unless the wheel rotational acceleration is close to 0, the road surface friction coefficient can be estimated only intermittently. Therefore, when the road surface friction coefficient changes, accurate estimation is possible. The value may not be obtained.
In the apparatus b), for example, when the left and right wheels are at the same wheel speed while traveling straight, the estimated road friction coefficient cannot be obtained, and therefore the estimated road friction coefficient can only be obtained intermittently.
In the apparatus c), since the apparatus for estimating the road surface friction coefficient does not operate when going straight ahead, the road surface friction coefficient estimated value can be obtained only intermittently.
[0004]
In the apparatus d), it is necessary to detect the total acceleration. However, since it is difficult to accurately detect the total acceleration without being affected by pitching or rolling, the estimated value of the road surface friction coefficient based on the total acceleration is used. Tends to be inaccurate.
In the apparatus e), a special sensor or the like is required to detect the stress of the suspension, so that it is difficult to realize at a low cost. In addition, the stress detection method must be changed according to the type of suspension.
[0005]
The apparatus disclosed in Japanese Patent Laid-Open No. 5-208670 of f) discloses an algorithm for measuring the slipperiness of the road surface by comparing the reference speeds V _sl , V _{sm and the} like with the wheel speed. Is an algorithm that classifies road surfaces in stages such as “slippery road surface”, “medium road surface”, and “slippery road surface”, and does not estimate the road surface friction coefficient itself as in the present invention. Further, the method disclosed in Japanese Patent Application Laid-Open No. 9-263226 in the above f) discloses a technique for classifying the slipperiness of the road surface from the relationship between the time when the brake fluid pressure is applied and the wheel speed. A technique for estimating the friction coefficient value itself is not disclosed.
[0006]
An object of this invention is to provide the road surface friction coefficient estimation apparatus which can estimate the road surface friction coefficient value itself accurately.
[0007]
[Means for Solving the Problems]
For the above purpose, a road surface friction coefficient estimating apparatus according to claim 1 of the present invention includes a wheel speed detecting means for detecting a wheel speed, a vehicle speed detecting means for detecting a vehicle speed, a braking force and a driving force of the wheel. A braking / driving force detecting means for detecting, a slip ratio calculating means for calculating the slip ratio of a wheel by inputting the wheel speed and the vehicle speed, and a slip coefficient of friction generated between the road surface and the tire by inputting the slip ratio. A pseudo μ-s function generating means for generating a pseudo μ-s function which is a function similar to the μ-s function expressed as a function of the rate, the wheel speed, the braking / driving force, and the estimated value of the road surface friction coefficient A wheel speed estimation error calculating means for calculating a wheel speed estimation value according to a road surface friction coefficient estimated value and calculating a wheel speed estimation error by subtracting the wheel speed estimated value from the wheel speed; and the wheel speed estimation error and Calibration parameter calculation means for calculating a calibration parameter for calibrating the pseudo μ-s function by inputting a similar μ-s function, and the pseudo μ-s function by inputting the pseudo μ-s function and the calibration parameter Is calibrated with the calibration parameters, and the calibration value is output as the road surface friction coefficient estimation value. The pseudo μ-s function is expressed by the following equation:

Where _xw is a wheel speed and _xv is a function of a slip ratio λ between the road surface and the tire expressed by a vehicle speed, and the slip ratio λ is a predetermined negative constant −λ ₁ or more and a predetermined positive number. monotonously increasing when less than λ ₂ , and monotonously decreasing when the slip ratio λ is less than a predetermined negative constant −λ ₁ or more than a predetermined positive number λ ₂ ,
When the pseudo μ-s function is μ _qu , μ _qu can be expressed by the following expression: μ _qu = D sin [Ctan ⁻¹ (Bφ)] (2)
φ = 100 (1-E) λ + (E / B) tan (100Bλ) (3)
However, B, C, D, and E are calculated by a magic formula represented by constants determined by tire characteristics.
[0008]
A road surface friction coefficient estimating device according to claim 2 of the present invention includes a wheel speed detecting means for detecting a wheel speed, a vehicle speed detecting means for detecting a vehicle speed, and a braking / driving force for detecting a braking force and a driving force of the wheel. Detection means, slip ratio calculation means for calculating the wheel slip ratio by inputting the wheel speed and the vehicle body speed, and a friction coefficient generated between the road surface and the tire by inputting the slip ratio as a function of the slip ratio A pseudo μ-s function generating means for generating a pseudo μ-s function which is a function similar to the μ-s function, the wheel speed, the braking / driving force, and the road surface friction coefficient estimated value are inputted, and the road surface friction coefficient estimated value is input. A wheel speed estimation error calculating means for calculating a wheel speed estimation error by calculating a wheel speed estimation value corresponding to the wheel speed, subtracting the wheel speed estimation value from the wheel speed, and the wheel speed estimation error and the pseudo μ-s function. The Calibration parameter calculation means for calculating a calibration parameter for calibrating the pseudo μ-s function, and the pseudo μ-s function and the calibration parameter are input to calibrate the pseudo μ-s function with the calibration parameter. The wheel speed estimation error calculating means includes a monotonous non-decreasing function of the wheel speed estimation error, braking / driving force, and road surface friction coefficient estimated value. (Where the coupling coefficient for the estimated road friction coefficient is the wheel load divided by the wheel inertia moment, and the braking / driving force coupling constant is the reciprocal of the wheel inertia moment, the wheel speed estimation error or the wheel speed estimation error. (The positive and negative sign coupling constant is a predetermined negative value that is determined as appropriate)
Is the amount of change in the estimated wheel speed value.
[0009]
The road surface friction coefficient estimating device according to claim 3 of the present invention is such that the monotonous non-decreasing function of the wheel speed estimation error in the road surface friction coefficient estimating device according to claim 2 is a wheel speed estimation error and a predetermined gain that is appropriately determined. Or a product of a positive or negative sign of a wheel speed estimation error and a predetermined gain determined as appropriate.
[0010]
A road surface friction coefficient estimating device according to a fourth aspect of the present invention includes a wheel speed detecting unit that detects a wheel speed, a vehicle body speed detecting unit that detects a vehicle body speed, and a braking / driving force that detects a braking force and a driving force of the wheel. Detection means, slip ratio calculation means for calculating the wheel slip ratio by inputting the wheel speed and the vehicle body speed, and a friction coefficient generated between the road surface and the tire by inputting the slip ratio as a function of the slip ratio A pseudo μ-s function generating means for generating a pseudo μ-s function which is a function similar to the μ-s function, the wheel speed, the braking / driving force, and the road surface friction coefficient estimated value are inputted, and the road surface friction coefficient estimated value is input. A wheel speed estimation error calculating means for calculating a wheel speed estimation error by calculating a wheel speed estimation value corresponding to the wheel speed, subtracting the wheel speed estimation value from the wheel speed, and the wheel speed estimation error and the pseudo μ-s function. The Calibration parameter calculation means for calculating a calibration parameter for calibrating the pseudo μ-s function, and the pseudo μ-s function and the calibration parameter are input to calibrate the pseudo μ-s function with the calibration parameter. And a calibration means for outputting the calibration value as the road surface friction coefficient estimation value. The calibration parameter calculation means calculates the rate of change of the calibration parameter as a wheel estimation error, a pseudo μ-s function value, and a positive constant. It is characterized by calculating with a product.
[0011]
【The invention's effect】
According to the first aspect of the present invention, a function similar to the μ-s function representing the friction coefficient generated between the road surface and the tire as a function of the slip ratio based on the slip ratio of the wheel calculated from the wheel speed and the vehicle body speed. Is generated, a wheel speed estimation value and a wheel speed estimation error corresponding to the road surface friction coefficient estimation value are calculated based on the wheel speed, braking / driving force, and road surface friction coefficient estimation value, and the wheel A calibration parameter used to calibrate the pseudo μ-s function is calculated based on the speed estimation error and the pseudo μ-s function, and a calibration value obtained by calibrating the pseudo μ-s function with the calibration parameter is used as the road surface friction coefficient estimation value. Therefore, the estimated value of the road surface friction coefficient quickly converges to the true value of the road surface friction coefficient. Therefore, it is possible to provide a road surface friction coefficient estimation device that can accurately estimate the road surface friction coefficient value itself.
According to claim 1 of the present invention, the pseudo μ-s function is obtained when the slip ratio λ between the road surface and the tire is equal to or larger than a predetermined negative constant −λ ₁ and smaller than a predetermined positive number λ _2. The pseudo μ-s function is generated between the road surface and the tire because it increases monotonously and decreases monotonously when the slip ratio λ is less than a predetermined negative constant −λ ₁ or a predetermined positive number λ ₂ or more. This is a function similar to the μ-s function that represents the friction coefficient as a function of the slip ratio. Therefore, by calibrating the pseudo μ-s function with the calibration parameters, a function in which the road surface friction coefficient is expressed by the slip ratio can be obtained.
Further, according to claim 1 of the present invention, the pseudo μ-s function μ _qu is calculated by the magic formula expressed by the above equations (2) and (3), and therefore the pseudo μ-s function is calculated by the road surface. This function is suitable for estimating the friction coefficient.
[0012]
According to the second aspect of the present invention, a function similar to the μ-s function representing the friction coefficient generated between the road surface and the tire as a function of the slip ratio based on the slip ratio of the wheel calculated from the wheel speed and the vehicle body speed. Is generated, a wheel speed estimation value and a wheel speed estimation error corresponding to the road surface friction coefficient estimation value are calculated based on the wheel speed, braking / driving force, and road surface friction coefficient estimation value, and the wheel A calibration parameter used to calibrate the pseudo μ-s function is calculated based on the speed estimation error and the pseudo μ-s function, and a calibration value obtained by calibrating the pseudo μ-s function with the calibration parameter is used as the road surface friction coefficient estimation value. When output, a linear combination of the monotonous non-decreasing function of the wheel speed estimation error, braking / driving force and road friction coefficient estimated value is used as the amount of change in the wheel speed estimated value. The vibration of the value can be suppressed.
[0013]
According to a third aspect of the present invention, the monotonous non-decreasing function of the wheel speed estimation error is a product of the wheel speed estimation error and a predetermined gain that is determined as appropriate, or a positive / negative sign of the wheel speed estimation error. Since it is determined by the product with the gain, the monotonous non-decreasing function of the wheel speed estimation error is a function suitable for the estimation of the road surface friction coefficient.
[0014]
According to the fourth aspect of the present invention, a function similar to the μ-s function representing the friction coefficient generated between the road surface and the tire as a function of the slip ratio based on the slip ratio of the wheel calculated from the wheel speed and the vehicle body speed. Is generated, a wheel speed estimation value and a wheel speed estimation error corresponding to the road surface friction coefficient estimation value are calculated based on the wheel speed, braking / driving force, and road surface friction coefficient estimation value, and the wheel A calibration parameter used to calibrate the pseudo μ-s function is calculated based on the speed estimation error and the pseudo μ-s function, and a calibration value obtained by calibrating the pseudo μ-s function with the calibration parameter is used as the road surface friction coefficient estimation value. At the time of output, the change rate of the calibration parameter is calculated by the product of the wheel estimation error, the pseudo μ-s function value, and a positive constant. It becomes suitable for the calculation of the over data.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a diagram showing a basic configuration of a road surface friction coefficient estimating apparatus according to the present invention. As shown in FIG. 1, a road surface friction coefficient estimating apparatus according to the present invention includes a wheel speed sensor 1 as a wheel speed detecting means, a vehicle speed sensor 2 as a vehicle speed detecting means, a braking / driving torque detecting means 3, and a slip ratio calculating means. 4. Pseudo μ-s function generating means 5, wheel speed estimation error calculating means 6, calibration parameter calculating means 7 and calibration means 8.
[0022]
Wheel speed sensor 1 detects the wheel speed (rotational speed of the wheel) x _w. As the wheel speed sensor 1, a device that periodically measures pulses generated by an tacho generator or a rotary incremental encoder can be used.
Vehicle speed sensor 2 detects a vehicle speed x _v. As the vehicle body sensor 2, one that detects the rotational speed of the non-driving wheel or the non-braking wheel and uses the detected value as the vehicle body speed can be used. Instead of the above, an optical ground speedometer using a spatial filter, a Doppler radar, or the like may be used.
[0023]
The braking / driving force detecting means 3 detects brake torque and driving torque (hereinafter referred to as braking / driving torque) u, which is the braking / driving force of the wheels. With respect to this braking / driving torque, the brake fluid pressure can be regarded as an approximate value corresponding to the braking / driving torque during braking and the engine torque during driving. For example, when the engine torque estimated value q _e is input from the engine control device and the brake fluid pressure p _b is input from a pressure sensor attached to the brake cylinder, the following equation u = c _e q _e + c _b p _b (4)
Here, c _e is a conversion coefficient from engine torque to braking / driving torque, and c _b is a braking / driving torque u based on a conversion coefficient from brake fluid pressure to braking / driving torque.
[0024]
Slip rate calculating means 4 is inputted to the wheel speed x _w and the vehicle body speed x _v, the slip ratio of the wheel lambda, following equation 3]

Calculate by The vehicle body speed x _v can be obtained from the vehicle body translation speed V _v and the wheel radius r _w by x _v ≈V _v / r _w .
[0025]
The pseudo μ-s function generating means 5 receives the slip ratio λ and is a function similar to the μ-s function that represents the friction coefficient generated between the road surface and the tire as a function of the slip ratio. _qu , the following formula:

Calculate by The pseudo μ-s function μ _qu has a shape in which a linear function having a negative slope −k ₁ is connected to both ends of a linear function having a positive slope k ₂ as shown in FIG. The shape is similar to the μ-s function as illustrated in FIG.
[0026]
The pseudo μ-s function generation means 5 is replaced with the magic formula μ _qu = Dsin [Ctan ⁻¹ (Bφ)] expressed by the following equation instead of the above equation (5) (2)
φ = 100 (1-E) λ + (E / B) tan (100Bλ) (3)
The pseudo μ-s function μ _qu can also be calculated. It is known that when an appropriate coefficient is given to B, C, D, and E in this magic formula, an actual μ-s function can be approximated with high accuracy (reference: H. B. Pacejka, Tire models for vehicle dynamics analysis, Supplement to Vehicle System Dynamics, vol. 21, (1991)). In this embodiment, for example, B = 0.1, C = 1.55, D = 1, and E = 0.432 are given. In this case, the pseudo μ-s function μ _qu has the shape shown in FIG. Become.
[0027]
The wheel speed estimation error calculation means 6 receives the braking / driving torque u, the wheel speed _xw, and a road surface friction coefficient estimated value μ _pr to be described later, and the following equation:

Where W is a wheel load, I is a wheel moment of inertia, h ₁ and h ₂ are positive constants determined by the designer as appropriate, and sgn (x _wpr −x _w ) is a sign of (x _wpr −x _w ) , -1, 0, 1 to _{obtain a} wheel speed estimated value x _wpr . Then, a wheel speed estimation error e is _{calculated from the} wheel speed estimated value x _wpr and the wheel speed x _w by the following equation e = x _w −x _wpr (7)
Calculate by
[0028]
The calibration parameter calculation means 7 is inputted with the wheel speed estimation error e and the pseudo μ-s function μ _qu , and the following equation:

However, c calculates a calibration parameter θ _pr for calibrating the pseudo μ-s function μ _qu with a positive constant appropriately determined by the designer.
[0029]
The calibration means 8 receives the pseudo μ-s function μ _qu and the calibration parameter θ _pr , and the following expression μ _pr = θ _pr μ _qu (9)
Is used to calibrate the pseudo μ-s function μ _qu with the calibration parameter θ _pr , and the calibration value is output as the road surface friction coefficient estimated value μ _pr .
[0030]
Next, the road surface friction coefficient estimation principle in this embodiment will be described.
The friction coefficient between the road surface and the tire varies depending on the slip ratio between the tire and the road surface, but the function μ (λ) representing the road surface friction coefficient μ as a function of the slip ratio λ is as shown in FIG. Regardless of the type (dry road, wet road, icing road), it has a similar shape. Therefore, when the unknown calibration parameter is θ, the function μ (λ) representing the road surface friction coefficient is approximately μ by the pseudo μ-s function μ _qu (λ) that is a known function and the unknown calibration parameter θ. (Λ) ≈θμ _qu (λ) (10)
And can be approximated.
[0031]
Therefore, in order to obtain the estimated value μ _pr of μ, using the estimated value θ _pr of θ, the following expression μ _pr = θ _pr μ _qu (11)
It is sufficient to perform the operation. Here, since the estimated value θ _pr is a parameter for calibrating the pseudo μ-s function μ _qu , the estimated value θ _pr itself is hereinafter referred to as a calibration parameter.
[0032]
The estimated road friction coefficient μ _pr is given by the following equation:

I think that it is expressed. In this case, μ takes a finite value, so
[Outside 1]

Is also a finite value. That is,
[Equation 8]

There is a positive constant γ _{0 such} that
[0033]
Now, the equation of motion of the wheel is
[Equation 9]

However, _xw can be expressed as wheel speed (wheel rotation speed), W as wheel load, u as braking / driving torque, and I as wheel inertia moment.
[0034]
Based on the equation (15), the calibration parameter θ _pr can be obtained by the following calculation.
[Expression 10]

here,

Where x _wpr is an estimated value of x _w , h ₁ , h ₂ and c are positive constants determined by the designer as appropriate, and h ₂ is
h ₂ > γ ₀ W / I (17)
However, γ ₀ is selected so as to satisfy a positive constant.
[0035]
The fact that the calibration parameter θ _pr converges to θ with time can be shown as follows.
Subtracting both sides of (6) from both sides of (15),
[Expression 11]

Get.
[0036]
On the other hand, in order to guarantee the convergence of the error e and (θ−θ _pr ) by a simple method, a function V having a positive value represented by the following equation is assumed:
V = (1/2) e ² + (c / 2) (θ−θ _pr ) ² (19)
Considering the time change, when θ is assumed to be a constant, the time change of V dV / dt is
[Expression 12]

Get.
[0037]
Here, if θ is changed according to the equation (8), the second term on the right side becomes 0, and h ₁ > 0, h ₂ > γ ₀ W / I. Therefore, the time change dV / dt of V is expressed by the following equation: [Formula 13]

Thus, the function V monotonously decreases and converges to 0. That is, since e and (θ−θ _pr ) also converge to 0, θ _pr converges to θ. Therefore, μ _pr = θ _pr μ _qu converges to μ.
[0038]
As is clear from the equation (21), h ₁ can be set to 0 and the term relating to h _{1 in the} equation (6) can be omitted. However, as h ₁ and h ₂ are increased, the estimation speed of the road surface friction coefficient is improved, and as h ₂ is increased,

Even if becomes larger, it becomes easier to estimate the correct road friction coefficient. However, when the value of h ₂ becomes excessive, case affected by the noise the noise on the detected value of the wheel speed x _w are mixed, estimates may vibrate. Therefore, in the present embodiment, sgn (x _wpr −x _w ) is substituted by the following monotonically non-decreasing continuous function f ₁ or f ₂ , which suppresses vibrations in the estimated wheel speed value. It is known that
_{f 1 (x wpr -x w)} : = | x wpr -x w | / (x wpr -x w + ε) (23)
Here, “: =” means “define the left side as the right side”, and ε is a positive constant.
[Expression 14]

Here, f ₀ and η ₀ are positive constants determined by the designer.
[0039]
According to the road surface friction coefficient estimating apparatus of the present embodiment, the following effects can be obtained.
(1) If slipping has occurred in any of the wheels, the road surface friction coefficient can be estimated.
(2) The road surface friction coefficient can be estimated quickly and accurately.
(3) The present invention can be applied to any type of wheel as long as the inertia moment I and wheel load W of the wheel are known.
(4) If the wheel speed and the vehicle body speed can be detected, a road surface friction coefficient estimating device can be configured.
(5) A large calibration parameter θ means that the road surface is hard to slip, and a small calibration parameter θ means that the road surface is slippery. Can be determined.
[0040]
【Example】
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 5 is a diagram showing a configuration when the road surface friction coefficient estimating apparatus according to the first embodiment of the present invention is mounted on a vehicle. In this embodiment, a brake fluid pressure sensor 12 and a wheel speed sensor 13 are provided in the vicinity of a tire (only one wheel is shown in FIG. 5), and these sensors 12 and 13 are connected to the computer unit 14. At the same time, an engine control device 16 and a vehicle speed sensor 17 for the engine 15 are connected to the computer unit 14. In this embodiment, a rotary encoder is used as the wheel speed sensor 13 and a ground vehicle speedometer is used as the vehicle body speed sensor 17. The detected value of the brake fluid pressure sensor 12 and the engine torque output from the engine control device 16 are used. The braking / driving torque is calculated based on the estimated value.
[0041]
The computer unit 14 performs a process of calculating the braking / driving torque from the brake fluid pressure sensor detection value and the engine torque estimation value, and also calculates the slip ratio calculation means, the pseudo μ-s function generation means, and the wheel speed estimation error calculation described above. The processing corresponding to the means, the calibration parameter calculation means, and the calibration means is performed. These processes are performed for each wheel speed _{x w} sufficiently short set sample time to changes (e.g. 10 [msec]).
[0042]
FIG. 6 is a flowchart showing a road surface friction coefficient estimation process by the computer unit in the first embodiment. In FIG. 6, first, at step 51, it is determined whether or not the sample time is reached, and the brake fluid pressure p _b is read from the brake fluid pressure sensor 12 at the next step 52 only when the sample time is reached. Then, in step 53, it reads the vehicle speed _{x v} from the vehicle speed sensor 17 reads in the wheel speed _{x w} from the wheel speed sensors 13.
[0043]
In the next step 55, the slip ratio λ is expressed by the following equation:

Calculate by
In the next step 56, the pseudo μ-s function μ _qu is expressed by the following expression: μ _qu = μ _qu (λ) (25)
Then, in step 57, the road surface friction coefficient estimated value μ _pr is calculated as follows: μ _pr = θ _pr μ _qu (9)
Calculate by
[0044]
In the next step 58, the following equation:

The wheel speed estimation error e is calculated by the following equation. Then, in step 59, the calibration parameter θ _pr is calculated by the following equation:

Calculate by
[0045]
FIGS. 7A to 7C are diagrams showing estimation results of the road surface friction coefficient according to the first embodiment. FIGS. 7A to 7C show the situation when the road surface friction coefficient is estimated when the automobile accelerates. That is, when the drive torque u increases, the slip ratio λ also increases accordingly, and the road surface friction coefficient μ and its estimated value μ _pr also change. In this embodiment, the road surface friction coefficient estimated value μ _pr is once set to the true value μ. As a result, the road surface friction coefficient estimated value μ _pr can be obtained with high accuracy even if the road surface friction coefficient μ changes. Further, the convergence of the road surface friction coefficient μ is very quick.
[0046]
The present embodiment is not limited to the above-described example, and various changes or modifications can be added. For example, in an automobile, usually four wheel speed detection values can be obtained, and from among the four wheel speed detection values, the smallest one during driving and the largest one during braking are selected as vehicle body speed detection values. If it substitutes, the estimation precision of a road surface friction coefficient will fall a little, but the apparatus structure can be simplified.
[0047]
In addition, when the absolute value of the slip ratio is small and therefore the absolute value of the frictional force itself is small, the calculation accuracy may be lowered. Therefore, the road friction coefficient may be estimated by weighting the value of θ when the absolute value of the slip ratio is large. That is, the road surface friction coefficient estimated value μ _pr is expressed by the following equation:

However, W (λ) may be calculated by an increasing function of λ.
[Brief description of the drawings]
FIG. 1 is a diagram showing a basic configuration of a road surface friction coefficient estimating apparatus according to the present invention.
FIG. 2 is a diagram illustrating a μ-s function representing a friction coefficient generated between a road surface and a tire as a function of a slip ratio.
FIG. 3 is a diagram illustrating an example of a pseudo μ-s function in the first embodiment of the invention.
FIG. 4 is a diagram illustrating another example of the pseudo μ-s function in the first embodiment of the present invention.
FIG. 5 is a diagram showing a configuration when the road surface friction coefficient estimating apparatus according to the first embodiment of the present invention is mounted on a vehicle.
FIG. 6 is a flowchart showing a road surface friction coefficient estimation process by a computer unit in the first embodiment.
FIGS. 7A to 7C are diagrams showing estimation results of a road surface friction coefficient according to the first embodiment. FIGS.
[Explanation of symbols]
1 Wheel speed sensor (wheel speed detection means)
2 Body speed sensor (body speed detection means)
3 Braking / driving torque detecting means (braking / driving force detecting means)
4 Slip rate calculation means 5 Pseudo μ-s function generation means 6 Wheel speed estimation error calculation means 7 Calibration parameter calculation means 8 Calibration means

Claims (4)

Wheel speed detecting means for detecting the wheel speed;
Vehicle speed detection means for detecting the vehicle speed;
Braking / driving force detecting means for detecting the braking force and driving force of the wheels;
Slip ratio calculating means for calculating the slip ratio of the wheel by inputting the wheel speed and the vehicle body speed;
Pseudo μ-s function generating means for generating a pseudo μ-s function that is a function similar to a μ-s function that represents the friction coefficient generated between the road surface and the tire as a function of the slip ratio when the slip ratio is input;
The wheel speed, braking / driving force, and road surface friction coefficient estimated value are inputted to calculate a wheel speed estimated value corresponding to the road surface friction coefficient estimated value, and the wheel speed estimated value is subtracted from the wheel speed to estimate the wheel speed. A wheel speed estimation error calculating means for calculating an error;
Calibration parameter calculation means for inputting the wheel speed estimation error and the pseudo μ-s function and calculating a calibration parameter for calibrating the pseudo μ-s function;
Calibration means for inputting the pseudo μ-s function and a calibration parameter, calibrating the pseudo μ-s function with the calibration parameter, and outputting the calibration value as the road friction coefficient estimation value;
The pseudo μ-s function is given by

Where _xw is a wheel speed and _xv is a function of a slip ratio λ between the road surface and the tire expressed by a vehicle speed, and the slip ratio λ is a predetermined negative constant −λ ₁ or more and a predetermined positive number. monotonously increasing when less than λ ₂ , and monotonously decreasing when the slip ratio λ is less than a predetermined negative constant −λ ₁ or a predetermined positive number λ ₂ or more,
When the pseudo μ-s function is μ _qu , μ _qu can be expressed by the following expression: μ _qu = D sin [Ctan ⁻¹ (Bφ)] (2)
φ = 100 (1-E) λ + (E / B) tan (100Bλ) (3)
However, B, C, D, and E are calculated by the magic formula represented by the constant determined by the tire characteristic, The road surface friction coefficient estimation apparatus characterized by the above-mentioned. Wheel speed detecting means for detecting the wheel speed;
Vehicle speed detection means for detecting the vehicle speed;
Braking / driving force detecting means for detecting the braking force and driving force of the wheels;
Slip ratio calculating means for calculating the slip ratio of the wheel by inputting the wheel speed and the vehicle body speed;
Pseudo μ-s function generating means for generating a pseudo μ-s function that is a function similar to a μ-s function that represents the friction coefficient generated between the road surface and the tire as a function of the slip ratio when the slip ratio is input;
The wheel speed, braking / driving force, and road surface friction coefficient estimated value are inputted to calculate a wheel speed estimated value corresponding to the road surface friction coefficient estimated value, and the wheel speed estimated value is subtracted from the wheel speed to estimate the wheel speed. A wheel speed estimation error calculating means for calculating an error;
Calibration parameter calculation means for inputting the wheel speed estimation error and the pseudo μ-s function and calculating a calibration parameter for calibrating the pseudo μ-s function;
Calibration means for inputting the pseudo μ-s function and a calibration parameter, calibrating the pseudo μ-s function with the calibration parameter, and outputting the calibration value as the road friction coefficient estimation value;
The wheel speed estimation error calculating means is
A linear combination of a monotonic non-decreasing function of wheel speed estimation error, braking / driving force, and estimated road friction coefficient (where the estimated coefficient of road friction coefficient is the wheel load divided by the wheel inertia moment, braking / driving force The coupling constant is the reciprocal of the wheel inertia moment, the wheel speed estimation error or the coupling constant of the sign of the wheel speed estimation error is a predetermined negative value that is determined as appropriate)
Is the amount of change in the estimated wheel speed value. In the road surface friction coefficient estimating device according to claim 2,
The monotonous non-decreasing function of the wheel speed estimation error is a product of a wheel speed estimation error and a predetermined gain that is determined as appropriate, or a product of a positive / negative sign of the wheel speed estimation error and a predetermined gain that is determined as appropriate. Road friction coefficient estimation device. Wheel speed detecting means for detecting the wheel speed;
Vehicle speed detection means for detecting the vehicle speed;
Braking / driving force detecting means for detecting the braking force and driving force of the wheels;
Slip ratio calculating means for calculating the slip ratio of the wheel by inputting the wheel speed and the vehicle body speed;
A pseudo μ-s function generating means for generating a pseudo μ-s function that is a function similar to a μ-s function that represents the coefficient of friction generated between the road surface and the tire as a function of the slip ratio when the slip ratio is input;
The wheel speed, braking / driving force and road surface friction coefficient estimated value are input to calculate a wheel speed estimated value corresponding to the road surface friction coefficient estimated value, and the wheel speed estimated value is subtracted from the wheel speed to estimate the wheel speed. A wheel speed estimation error calculating means for calculating an error;
Calibration parameter calculation means for inputting the wheel speed estimation error and the pseudo μ-s function and calculating a calibration parameter for calibrating the pseudo μ-s function;
Calibration means for inputting the pseudo μ-s function and a calibration parameter, calibrating the pseudo μ-s function with the calibration parameter, and outputting the calibration value as the road friction coefficient estimation value;
The calibration parameter calculation means calculates a change rate of a calibration parameter by a product of a wheel estimation error, a pseudo μ-s function value, and a positive constant.

Images