JP3751923B2 - Vehicle evaluation device - Google Patents

Description

【外１】

【数２】

ここで、ｍ₀は操舵トルク振幅、ωは操舵トルクの角速度、ｔは時間を表す。
【００２７】
次に、次式の無次元化パラメータを導入する。
【外２】

また、図７にステアリング振動モデルを示す。図は、ステアリングホイール４の操舵（操舵角θ、振動数ω）を、ばね定数Ｋ（θ，ω）及び減衰係数Ｃ（θ，ω）をもって表現することができることを示すものである。
【００２８】
上記無次元化パラメータの導入によって、以下のように（２）式を無次元化することができる。
【００２９】
【数３】

【数４】

【数５】

（３）〜（５）式を（２）式に代入すると以下の式を得る。
【００３０】
【数６】

【外３】

【数７】

【数８】

（７）式を（９）式で表すことができる。
【００３１】
【数９】

減衰項とばね剛性（周期）の関係を知るために共振曲線を求める場合があるが、ここでも同様に、特解のみの性質に注目して、（９）式の定常解はφ₀を角度、Ψ₀を位相差として次式で表すことができる。
【００３２】
【数１０】

本論の場合、復元力は非線形であるが、その定常解を（１０）式の形で表すと、（１０）式の導関数は次式となる。
【００３３】
【数１１】

ここで、φ₀、Ψ₀は時刻τとともにゆっくり変化すると考え、上式の第２、第３項の和を０とおく。（ただし、ξ＝ητ＋Ψ₀、'の記号はτに関する１階微分を示す。）
【数１２】

（１１）、（１２）式の関係を用いると（９）式は次式で表せる。
【００３４】
【数１３】

(12)式×（−ηcosξ）＋(5-13)式×sinξおよび(5-12)式×（ηsinξ）＋(5-13)式×cosξをつくれば、それぞれ次式を得る。
【００３５】
【数１４】

【数１５】

ところで、φ₀、Ψ₀は時刻τとともにゆっくり変化するSlowly varying parameterであり、φ₀'、Ψ₀'が０であれば、φ₀、Ψ₀は一定値となるが、厳密に０にすることは出来ない。φ₀'、Ψ₀'の項は、（１４）、（１５）式を０〜２πまでの一周期について積分して平均したものを０と考えると次式を得る。
【００３６】
【数１６】

【数１７】

（１６）及び（１７）式はそれぞれ次式となる。
【００３７】
【数１８】

【数１９】

ただし、
【００３８】
【数２０】

【数２１】

（１８）及び（１９）式からφ₀、Ψ₀を整理すれば、それぞれ次式を得る。
【００３９】
【数２２】

【数２３】

このように、非線形振動方程式（９）式から、（２２）及び（２３）式の関係がＳ（φ₀,η）、Ｃ（φ₀,η）を用いて求められる。
【００４０】
一方、（５）〜（９）式の履歴振動系を次式の等価な線形振動系に置換することを考える。
【００４１】
【数２４】

ここで、Ｈ_eqは等価減衰係数、Ｋ_eqは等価ばね定数を表す。
【００４２】
上式は線形であるから、解を簡単に求めることができる。（２２）式と同じ形式で表すと次のようになる。
【００４３】
【数２５】

また、（２４）式が（９）式の履歴系と振動特性が一致するための条件は、（２２）式と（２５）式とが恒等式的に等しいことであり、そのためには次式が成立しなければならない。
【００４４】
【数２６】

【数２７】

このことは、Ｓ（φ₀,η）が減衰特性に、Ｃ（φ₀,η）がばね特性に直接関係していることを表している。また、Ｓ（φ₀,η）、Ｃ（φ₀,η）が求まれば、等価減衰係数Ｈ_eq、等価ばね定数Ｋ_eqが決定されることになる。
【００４５】
ところで、Ｓ（φ₀,η）、Ｃ（φ₀,η）と履歴曲線の形状との関係について調べる必要がある。（２０）式においてφ＝φ₀cosξとおけば、
【００４６】
【数２８】

ξ＝０，２πで、φ＝φ₀であるから、
【００４７】
【数２９】

ここで、Ａ（φ₀,η）は周波数ηの時の角度φ₀に対応する履歴曲線の囲む面積である。
【００４８】
（２９）式は、Ｓ（φ₀,η）が履歴曲線の囲む面積に比例し、角度の２乗に反比例することを表している。また（２９）式は、履歴曲線の囲む面積さえ等しければ形状によらずＳ（φ₀,η）が等しくなることを示している。
【００４９】
一方、（２１）式を次のように変形できる。
【００５０】
【数３０】

ここで、φ₀cosξ＝φとおけば、cotξを次式で表すことができる。
【００５１】
【数３１】

（３１）式にあっては、ξ＝０〜２πの間のおいて、ξ＝０,π／２,πでは順次φ＝φ₀,０,−φ₀となることから、正弦または余弦入力に対する定常振動状態での範囲はＭ（φ,η）の減力線に対応し、（３１）式の符号は＋を取る。また、ξ＝π,３π／２,２πでは順次φ＝−φ₀,０,φ₀となることから、Ｍ（φ,η）の加力線に対応し、（３１）式の符号は−になる（表１参照）。
【００５２】
【表１】

このことを用いて、（３０）式は次のように変形できる。
【００５３】
【数３２】

（３２）式において、
【００５４】
【数３３】

であり、Ｒ（φ,η）を和曲線と呼ぶ。また、定常解を対象とする履歴曲線は原点対称であることから（３２）式は次のように簡略化できる。
【００５５】
【数３４】

上記各式中のＰ（φ）は定関数であることから、Ｒ（φ,η）、すなわち周波数ηの時の加力線と減力線とを同じ角度レベルで加えた和曲線さえ同じであれば、履歴曲線の形状によらずＣ（φ₀,η）は等しくなる。ここで、Ａ（φ₀,η）、Ｒ（φ₀,η）を算定するための履歴曲線の基本モデルとして、図８に示されたべき関数型復元力モデルを適用する。その基本式は次式で表せる。
【００５６】
【数３５】

ただし、α、ｋはφ₀とηとの関数である。
【００５７】
図８に示される復元力モデルにおいて、その履歴曲線の囲む面積Ａ（φ₀,η）を次式のように計算できる。
【００５８】
【数３６】

（３５）式の骨曲線と（３６）式の面積を求める式とにより、α、ｋは次の各式となる。
【００５９】
【数３７】

【数３８】

また、（３６）式を（２９）式に代入すると、
【００６０】
【数３９】

となる。また、和曲線Ｒ（φ,η）は、
【００６１】
【数４０】

であるから、（３４）式からＣ（φ₀,η）を次式のように求めることができる。
【外４】

【数４１】

なお、（４１）式中でΓはΓ関数を示す。
【００６２】
（２６）式と（２７）式とにそれぞれ（３９）式と（４１）式とを代入すれば、等価ばね定数Ｋ_eq、等価減衰係数Ｈ_eqが次のように求められる。
【００６３】
【数４２】

【数４３】

すなわち、べき関数型復元力モデルのα、ｋが求められれば、等価ばね定数Ｋ_eq、等価減衰係数Ｈ_eqが定まることになる。
【００６４】
次に、上記べき関数型復元力モデルの形状パラメータα、ｋの求め方について示す。ここでは、履歴曲線の基本ループをべき関数型復元力モデルへ置換する方法を述べる。
【００６５】
振動特性を反映させる条件は以下の２つである。
（Ｉ） 動的加力試験から得られる履歴曲線の囲む面積と、べき関数型復元力モデルの履歴曲線の囲む面積を等しくする（減衰量の一致）。
（ＩＩ） 動的加力試験から得られる履歴曲線の各頂点を結んで得られる骨曲線とべき関数型復元力モデルの骨曲線とを一致させる。すなわち、各々の各変位レベルにおいて動的加力試験と復元力モデルの頂点を一致させる（剛性量の一致）。
【００６６】
今、動的加力試験から得られた履歴ループにおいて、その履歴曲線の囲む面積および骨曲線を操舵角φ₀と外力の周波数ηとによりそれぞれＧ₀（φ₀,η）とＦ₀（φ₀,η）とに関数化できたものとする。ただし、Ｇ₀（φ₀,η）及びＦ₀（φ₀,η）はθ_s、ｍ_sで無次元化されているものとする。
【００６７】
（Ｉ）の条件は（３４）式を用いて次式のように表せる。
【００６８】
【数４４】

（ＩＩ）の条件は、（３３）式の骨曲線の基本式から次式のように表せる。
【００６９】
【数４５】

この（４５）式を（４４）式に代入してｋを消去すると、次式になる。
【００７０】
【数４６】

（４６）式からαは次式のように求められる。
【００７１】
【数４７】

また、ｋは（４５）式から次式のように求められる。
【００７２】
【数４８】

【００７３】
（４７）及び（４８）式で表せるα（φ₀,η）及びｋ（φ₀,η）を用いれば、動的加力試験とべき関数型復元力モデルとのそれぞれの履歴曲線の囲む面積及び骨曲線を一致させることができる。すなわち、Ｇ₀、Ｆ₀を操舵角φ₀、外力の周波数ηで関数化できれば、α（φ₀,η）及びｋ（φ₀,η）を用いて、べき関数型復元力モデルへ置換できることになる。このα（φ₀,η）及びｋ（φ₀,η）を（４２）式と（４３）式とに代入して等価ばね定数Ｋ_eqと等価減衰係数Ｈ_eqとを求めることてができ、以下の（４９）及び（５０）式からステアリング系の操舵角θおよび操舵速度（周波数）ωに依存するばね定数Ｋ（θ₀,ω₀）と減衰係数Ｃ（θ₀,ω₀）とを求めることができる。
【数４９】

【数５０】

次に、実車を用いて確認された測定結果について示す。まず、操舵性の良いＡ車と悪いとされるＢ車とを用いて、平坦なドライアスファルト路面を例えば車速１４０ｋｍ／ｈ一定で最大操舵角及び操舵速度（周波数）を段階的に変化させて正弦波スラローム走行を行い、操舵周波数毎の操舵角−操舵トルクヒステリシスカーブを測定した。例えばＡ車が図３（ｂ）に示されるようになり、Ｂ車が図３（ａ）に示されるようになる。具体的には、操舵周波数が０．２５、０．５、０．７５、１．０Ｈｚ毎で、操舵角が５、１０、１５、２０度の時の操舵角操舵トルクヒステリシスカーブを測定し、ヒステリシスカーブから骨曲線Ｍ₀（φ,η）と、ヒステリシスカーブの囲む面積Ｓ₀（φ,η）を求めた。この骨曲線が動ばね定数に対応し、面積が減衰係数に対応し得る。
【００７４】
求めた骨曲線Ｍ₀（φ,η）と面積Ｓ₀（φ,η）とを用いて、（４７）及び（４８）式からα及びｋの各値を求めた。一例として操舵周波数が０．５Ｈｚの時の操舵角に対するα及びｋの各値について、Ａ車を丸印で、Ｂ車をバツ印で示したものを図９に示す。図９（ａ）は操舵角に対するαであり、図９（ｂ）は操舵角に対するｋである。
【００７５】
また、上記したように求めたα、ｋを用いて（２６）及び（５０）式から操舵ばね定数を求めた。図１０に、操舵周波数０．２５Ｈｚの操舵角に対する操舵ばね定数のＡ車（丸印）とＢ車（バツ印）との比較を示す。図に示されるように、操舵性の良いＡ車に対して悪いＢ車は、操舵角が小さい領域で操舵速度が緩やかな状況（操舵周波数が低い状況）において操舵ばね定数Ｋが低くなっていることが分かる。これは、この領域における操舵時での手応えの無さを定量的に表現しており、Ｂ車の操舵性を悪くしている大きな要因であると言える。
【００７６】
同様にα、ｋを用いて（２７）及び（５１）式から操舵減衰係数を求めた。図１１に操舵周波数０．２５Ｈｚの操舵角に対する操舵減衰係数のＡ車（丸印）とＢ車（バツ印）との比較を示す。図に示されるように、操舵性が良いＡ車に対しＢ車は、すべての領域において操舵減衰係数が小さいことがわかる。これは、操舵時の位置決め性の無さを含めた広義な意味での安定性の無さを表現しており、これについてもＢ車の操舵性を悪くしている大きな要因であると言える。なお、操舵周波数０．５、０．７５、１．０Ｈｚについても同様であった。
【００７７】
このように、実車走行時における転舵周波数毎の操舵角‐操舵トルクのヒステリシスカーブの測定値からステアリング系の非線形振動特性（ばね定数、減衰係数）を求める手法として、べき関数型復元力モデルを用いた等価線形系解析手法を用いて操舵性の良い車と悪い車を解析した。その結果、転舵周波数および操舵角に大きく依存するステアリング系の非線形振動特性が求められることが分かった。また、本解析手法を用いて解析した結果から、従来手法では困難であった操舵力特性の非線形性を定量的に把握することが可能になり、操舵性の違う車両特性を定量的且つ明確に表現することができた。以上より、操舵性を定量的に解析する手法として新たな解析手法が提案できた。
【００７８】
なお、本図示例では操舵性の車両評価装置について示したが、本発明が適用される車両評価装置にあっては、操舵性に限られるものではなく、動ばねとみなして数値化できる部位に適用可能であり、ブレーキの評価や、サスペンションの評価や、加速性能の評価など車両全体に適用可能である。特に、非線形特性を示すものに有効である。
【００７９】
【発明の効果】
このように本発明によれば、入力の変化値とその入力に対応して変化する入力対応変化値とによりヒステリシスカーブが作成され、そのヒステリシスカーブの特性を動ばね定数と減衰係数との２つの指標値で表現できるため、強い非線形性を示すような車両に対しても十分な車両評価を行うことができる。そのため、ドライバーがステアリングを操作する時の操舵角度や操舵速度に対する操舵トルク（ステアリングを介して手に伝わってくる反力）やヨーレイトが非線形特性を有するものであっても、何ら問題なくその非線形特性を把握することができる。これにより、従来の評価方法では解析できなかった上記した非線形特性を有するものであっても定量的に把握することができ、自動車の重要な性能の１つである操舵性の車両評価を行うことができるようになった。
【００８０】
また、操舵角・操舵トルク・ヨーレイト・横加速度のいずれか２つを入力とその対応変化値として、ヒステリシスカーブを作成することにより、車両の評価として操舵性を評価することができるため、車両評価が困難であった操舵性を２つの指標値で評価でき、望ましい操舵フィーリングを容易に実現することができる。
【００８１】
また、例えば入力量依存特性の入力量を操舵角の大きさ（振幅）とし、入力周波数依存特性の入力周波数を操舵速度（周波数）とし、それら各入力値から動ばね定数と減衰係数とを算出することができる。これにより、実車走行時における操舵速度を段階的に変えた操舵力実験を行って得られた復元力特性から操舵角及び操舵速度に依存するステアリング系の非線形振動特性（ばね定数・減衰係数）を求めることが可能になり、その結果から車両特性の違いを明確に表現できる。
【図面の簡単な説明】
【図１】本発明が適用された車両の模式的平面図。
【図２】コントロールユニットによる制御の概略図。
【図３】操舵角−操舵トルクのヒステリシスカーブであって、（ａ）は調節対象車両の調節前であり、（ｂ）は目標車両であり、（ｃ）は調節後である。
【図４】操舵角−操舵トルクにおけるばね定数Ｋ及び減衰係数Ｃの評価指標マップ。
【図５】操舵角−ヨーレイトのヒステリシスカーブであって、（ａ）は調節対象車両の調節前であり、（ｂ）は目標車両であり、（ｃ）は調節後である。
【図６】操舵角−ヨーレイトにおけるばね定数Ｋ及び減衰係数Ｃの評価指標マップ。
【図７】ステアリング振動モデルを示す図。
【図８】べき関数型復元力モデルを示す図。
【図９】（ａ）は操舵性の良いＡ車と悪いＢ車とにおける操舵ばね定数を求めるための操舵角に対するαを示す図であり、（ｂ）は操舵角に対するｋを示す図。
【図１０】操舵性の良いＡ車と悪いＢ車とにおける操舵周波数０．２５Ｈｚでの操舵角に対するばね定数を示す図。
【図１１】操舵性の良いＡ車と悪いＢ車とにおける操舵周波数０．２５Ｈｚでの操舵角に対する減衰係数を示す図。
【符号の説明】
７ａ リサージュ作成部（ヒステリシスカーブ作成手段）
７ｂ 定数算出部（動ばね定数算出手段・減衰係数算出手段）[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a vehicle evaluation device relating to evaluation of traveling performance and the like.
[0002]
[Prior art]
Conventionally, information from various sensors (operation state detection means) is input to the control means, and the operation state is objectively evaluated and displayed, and a specific operation state set in advance is determined and alarmed. There is something that was made. (For example, refer to Patent Document 1).
[0003]
On the other hand, in order to evaluate vehicle characteristics (seasoning) such as EPS (electric steering system), brakes, suspension, acceleration performance, etc., for example, a prototype vehicle is evaluated and the result is reflected in the production of the actual vehicle. Can do. In conventional vehicle evaluation in such a vehicle evaluation apparatus, for example, a frequency response characteristic in a linear range is represented by a Bode diagram, and an evaluation method using this is used, or a partial hysteresis curve drawn by input / output data is used. There is a method of expressing with a simple gradient and hysteresis width.
[0004]
[Patent Document 1]
Japanese Unexamined Patent Publication No. 2000-247162 ([0025], FIG. 2)
[0005]
[Problems to be solved by the invention]
As a problem in the vehicle evaluation method described above, the vehicle characteristics often show strong nonlinearity, and the evaluation method using the Bode diagram cannot be sufficiently evaluated. In addition, the technique expressed by the slope and hysteresis width of each part of the hysteresis curve is complicated by a large number of evaluation index values. Further, when the input condition is changed, there is a problem that an enormous number of data is required to grasp the input amplitude dependency and the input frequency dependency, and the evaluation becomes complicated. For this reason, there is a case where it depends on intuition by a designer's skill, and there is a problem that vehicle evaluation cannot be easily performed.
[0006]
[Means for Solving the Problems]
In order to solve such a problem and to realize a vehicle evaluation apparatus that can easily grasp input amplitude dependency and input frequency dependency and can easily perform appropriate vehicle evaluation, in the present invention, , The vehicle's for any input Steerability Quantitative evaluation of A vehicle evaluation device for obtaining an assist gain and a damping gain as control parameters of a steering device of the vehicle, wherein any two of a steering angle, a steering torque, a yaw rate, and a lateral acceleration are obtained. A change value of the input and an input corresponding change value that changes corresponding to the input; do it Hysteresis curve The Hysteresis curve creating means for creating, dynamic spring constant calculating means for calculating a dynamic spring constant in vibration characteristics from the slope of the hysteresis curve, and damping coefficient calculating means for calculating a damping coefficient in vibration characteristics from the area of the hysteresis curve And the assist gain from each deviation obtained by comparing the target spring constant corresponding to the dynamic spring constant of the target vehicle and the target damping coefficient corresponding to the damping coefficient with the dynamic spring constant and the damping coefficient, respectively. And means for obtaining the damping gain It was supposed to be.
[0007]
According to this, since the characteristic of the hysteresis curve created by the input change value and the input corresponding change value that changes corresponding to the input can be expressed by two index values of the dynamic spring constant and the damping coefficient, It is possible to perform vehicle evaluation on a vehicle that exhibits a behavior with strong nonlinearity.
[0008]
The hysteresis curve creating means (7a) creates the hysteresis curve by using any two of a steering angle, a steering torque, a yaw rate, and a lateral acceleration as the input and input corresponding change values, and the vehicle According to the evaluation of the steering performance, it is possible to evaluate the steering performance, which is difficult to evaluate the vehicle, using a hysteresis curve created from two index values. Therefore, the desired steering performance can be easily realized by setting the control amount based on that.
[0009]
Further, by calculating the dynamic spring constant and the damping coefficient as a function of the steering angle and the steering speed, it is possible to grasp the input amount dependent characteristic and the input frequency dependent characteristic. For example, the input quantity dependent characteristic is for the magnitude (amplitude) of the steering angle, and the input frequency dependent characteristic is for the steering speed (frequency). From these input values, the dynamic spring constant and the damping coefficient are calculated. Can be calculated.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail based on specific examples shown in the accompanying drawings.
[0011]
FIG. 1 is a schematic plan view of a vehicle 1 to which the present invention is applied. In the illustrated example, vehicle evaluation of steering performance is performed. In the drawing, an EPS (electric steering device) 2 is used as a steering device for steering the front wheels FW. A steering shaft 3 is connected to the EPS 2. A steering angle sensor 5 for detecting a steering angle θ by steering of the steering wheel 4 and a steering torque sensor 6 for detecting a steering torque T are provided at appropriate positions of the steering shaft 3. Is provided. The detection signals of the sensors 5 and 6 are input to the control unit 7 that controls the EPS 2. The vehicle is provided with a yaw rate sensor 8 for detecting the yaw rate, and a detection signal of the yaw rate sensor 8 is input to the control unit 7.
[0012]
FIG. 2 shows an outline of control by the control unit 7. As an input signal for the change of the vehicle 1 to be evaluated, the steering angle θ or the steering speed by the steering of the steering wheel 4 can be used. As the steering speed, a steering angular speed ω that can be calculated from a time change of the steering angle θ is used. Further, the steering angle θ can be handled as an amplitude as an input quantity, and the steering angular velocity ω can be handled as an input frequency.
[0013]
A steering angle θ or a steering angular velocity ω as an input is generated according to the steering, and the behavior of the vehicle changes. For example, the steering angle θ is changed as the input change value, the change in the steering torque T or the yaw rate Y is detected as the input corresponding change value, and the steering angle θ and the detected value of the steering torque T or yaw rate Y are used as the hysteresis curve creating means. To the Lissajous creation unit 7a. In the Lissajous creation unit 7a, for example, a Lissajous waveform (hysteresis curve) of steering angle θ-steering torque T is created, and the waveform signal is output to a constant calculation unit 7b as a dynamic spring constant calculation unit and a damping coefficient calculation unit.
[0014]
Although two variables are used in creating the Lissajous waveform, the two variables in the case of evaluating the steering performance as in the illustrated example are not limited to the steering torque T or the yaw rate Y with respect to the steering angle θ. Any one of steering torque, yaw rate, and lateral acceleration may be used.
[0015]
The constant calculator 7b calculates a spring constant K and a damping coefficient C from the Lissajous waveform so that the vehicle evaluation can be handled as a vibration characteristic having the steering angle θ as an amplitude. The spring constant K and the damping coefficient C are respectively compared with the target spring constant Ko and the target damping coefficient Co by the comparator 7c, and each deviation (ΔK, ΔC) is calculated. As a means to find assist gain and damping gain Input to the controller 7d. controller 7d Then, the assist gain Ga and the damping gain Gd as EPS control parameters are optimized. Then, by controlling the EPS using the assist gain Ga and the damping gain Gd optimized by the controller 7d, it becomes possible to match the target value (Ko, Co).
[0016]
The Lissajous waveform created by the Lissajous creation unit 7a is as shown in FIG. 3, for example. FIG. 3 shows the relationship of the steering torque T with respect to the steering angle θ. For example, in a vehicle to be adjusted that controls the EPS 2 by the controller 7d in which an arbitrary assist gain Ga and damping gain Gd are set, a solid line represents a case where the steering angle θ is steered within a range of ± 5 degrees at a predetermined vehicle speed. When the case of ± 10 degrees is indicated by a broken line and the case of ± 15 degrees is indicated by an imaginary line, the relationship of the change of the steering torque T with respect to the change of the steering angle θ is shown in FIG.
[0017]
On the other hand, when the same Lissajous waveform of the target vehicle (ideal model vehicle) is shown in FIG. 3B from the steering property (steering feeling), the steering property of the vehicle is obtained. The target spring constant Ko and the target damping coefficient Co in the case where the vehicle evaluation is expressed by vibration characteristics are obtained in advance. These values can be obtained from the slope and area of the Lissajous waveform by an analysis method described later.
[0018]
Then, by obtaining the assist gain Ga and the damping gain Gd of the controller 5 from the deviations (ΔK, ΔC) with respect to the target spring constant Ko and the target damping coefficient Co by the method using the vehicle evaluation apparatus described above, FIG. It was possible to realize the steering performance with the Lissajous waveform shown in FIG. As a result, a vehicle having the same steering performance as that of the target vehicle can be obtained by controlling the EPS 2 and can be easily reflected in the actual vehicle.
[0019]
FIG. 4 shows an evaluation index map of the spring constant K and the damping coefficient C in the relationship between the steering angle and the steering torque. In the figure, the horizontal axis is the damping coefficient C, the vertical axis is the spring constant K, and the range 11 that is good in steering is shown by hatching. In the figure, the state of FIG. 3A is a cross mark, and the state of FIG. 3B is a square mark. The state shown in FIG. 3C after changing according to the target value entered the range 11 as indicated by a circle. When the spring constant K and the damping coefficient C are within the above range 11, good steering performance can be obtained. A map can be created based on the experimental results of an actual vehicle or a prototype vehicle with good steering performance.
[0020]
Note that the steering force characteristic can be evaluated from the relationship between the steering angle and the steering torque. That is, the damping coefficient C increases as the damping coefficient C increases, and the steering force increases as the spring constant K increases. In this case, by entering the range 11, an appropriate damping feeling and an appropriate steering force weight can be obtained.
[0021]
Whether the steering performance is good or not is not limited to the relationship of the change of the steering torque T with respect to the change of the steering angle θ, but is determined from the relationship of the change of the yaw rate Y to the change of the steering angle θ as shown in FIG. Anyway. 5, FIG. 5 (a) corresponds to FIG. 3 (a), FIG. 5 (b) corresponds to FIG. 3 (b), and FIG. 5 (c) corresponds to FIG. 3 (c). . Even in this way, it was possible to achieve the steering performance that resulted in a Lissajous waveform (FIG. 5C) substantially equivalent to the target Lissajous waveform of FIG.
[0022]
FIG. 6 shows an evaluation index map of the spring constant K and the damping coefficient C in the relationship between the steering angle and the yaw rate. In this case, the range 12 that provides good steering performance is indicated by hatching in the figure. As in FIG. 4, when the spring constant K and the damping coefficient C are within the above range 12, good steering performance can be obtained. This map can also be created based on the experimental results of actual vehicles and prototype vehicles with good steering performance.
[0023]
The vehicle response characteristics can be evaluated from the relationship between the steering angle and the yaw rate. This is a characteristic that the responsiveness decreases as the damping coefficient C increases (the yaw rate delay relative to the input steering angle increases), and the yaw gain increases as the spring constant K increases (becomes quick). . In this case, by entering the above range, the yaw rate delay is suppressed and the quick feeling is increased.
[0024]
Further, a Lissajous waveform similar to that described above is obtained from the relationship of the change in the yaw rate Y with respect to the change in the steering torque T, whereby the sense of unity with the vehicle during steering can be evaluated. As the damping coefficient C increases, the responsiveness decreases (the yaw rate delay with respect to the steering torque increases), and when the spring constant K increases, the yaw gain increases (the yaw rate is generated with a small steering torque). It is a characteristic.
[0025]
Next, a steering performance analysis method in the vehicle evaluation apparatus configured as described above will be described. First, the equation of motion of a spring mass system in which a forced external force generally acts is as follows.
[0026]
[Expression 1]

[Outside 1]

[Expression 2]

Where m ₀ Is the steering torque amplitude, ω is the angular velocity of the steering torque, and t is the time.
[0027]
Next, the dimensionless parameter of the following formula is introduced.
[Outside 2]

FIG. 7 shows a steering vibration model. The figure shows that the steering (steering angle θ, frequency ω) of the steering wheel 4 can be expressed by a spring constant K (θ, ω) and a damping coefficient C (θ, ω).
[0028]
By introducing the dimensionless parameter, the formula (2) can be dimensionless as follows.
[0029]
[Equation 3]

[Expression 4]

[Equation 5]

Substituting Equations (3) to (5) into Equation (2) gives the following equation.
[0030]
[Formula 6]

[Outside 3]

[Expression 7]

[Equation 8]

Expression (7) can be expressed by Expression (9).
[0031]
[Equation 9]

In order to know the relationship between the damping term and the spring stiffness (period), a resonance curve may be obtained. Similarly, paying attention to the property of only the special solution, the steady solution of the equation (9) is φ ₀ The angle, Ψ ₀ Can be expressed by the following equation as a phase difference.
[0032]
[Expression 10]

In this case, the restoring force is non-linear, but if the steady solution is expressed in the form of equation (10), the derivative of equation (10) becomes the following equation.
[0033]
## EQU11 ##

Where φ ₀ , Ψ ₀ Is considered to change slowly with time τ, and the sum of the second and third terms in the above equation is set to zero. (However, ξ = ητ + Ψ ₀ , 'Indicates the first derivative with respect to τ. )
[Expression 12]

Using the relationship between the expressions (11) and (12), the expression (9) can be expressed by the following expression.
[0034]
[Formula 13]

If formula (12) × (−ηcosξ) + (5-13) × sinξ and (5-12) × (ηsinξ) + (5-13) × cosξ are created, the following equations are obtained respectively.
[0035]
[Expression 14]

[Expression 15]

By the way, φ ₀ , Ψ ₀ Is a slowly varying parameter that changes slowly with time τ, and φ ₀ ', Ψ ₀ If 'is 0, φ ₀ , Ψ ₀ Is a constant value, but cannot be exactly zero. φ ₀ ', Ψ ₀ The term of 'is obtained by considering the equation (14) and (15), which is obtained by integrating and averaging over one period from 0 to 2π, as follows.
[0036]
[Expression 16]

[Expression 17]

Expressions (16) and (17) are respectively the following expressions.
[0037]
[Formula 18]

[Equation 19]

However,
[0038]
[Expression 20]

[Expression 21]

From Equations (18) and (19) ₀ , Ψ ₀ If we arrange, we get the following equations respectively.
[0039]
[Expression 22]

[Expression 23]

Thus, from the nonlinear vibration equation (9), the relationship between the equations (22) and (23) is S (φ ₀ , η), C (φ ₀ , η).
[0040]
On the other hand, consider replacing the hysteresis vibration system of the equations (5) to (9) with an equivalent linear vibration system of the following equation.
[0041]
[Expression 24]

Where H _eq Is the equivalent damping coefficient, K _eq Represents an equivalent spring constant.
[0042]
Since the above equation is linear, the solution can be easily obtained. When expressed in the same form as the equation (22), it is as follows.
[0043]
[Expression 25]

Further, the condition for the vibration characteristic to coincide with the hysteresis system of the expression (9) in the expression (24) is that the expression (22) and the expression (25) are identically equal. It must be approved.
[0044]
[Equation 26]

[Expression 27]

This means that S (φ ₀ , η) is the damping characteristic, C (φ ₀ , η) is directly related to the spring characteristics. Also, S (φ ₀ , η), C (φ ₀ , η), the equivalent damping coefficient H _eq , Equivalent spring constant K _eq Will be determined.
[0045]
By the way, S (φ ₀ , η), C (φ ₀ , η) and the shape of the history curve need to be investigated. In equation (20), φ = φ ₀ Cosξ
[0046]
[Expression 28]

ξ = 0, 2π, φ = φ ₀ Because
[0047]
[Expression 29]

Where A (φ ₀ , η) is the angle φ at the frequency η ₀ Is the area surrounded by the history curve corresponding to.
[0048]
Equation (29) is expressed as S (φ ₀ , η) is proportional to the area surrounded by the history curve and inversely proportional to the square of the angle. In addition, the expression (29) can be expressed as S (φ ₀ , η) are equal.
[0049]
On the other hand, the equation (21) can be modified as follows.
[0050]
[30]

Where φ ₀ If cosξ = φ, cotξ can be expressed by the following equation.
[0051]
[31]

In the equation (31), when ξ = 0, π / 2, π between ξ = 0 and 2π, φ = φ sequentially. ₀ , 0, -φ ₀ Therefore, the range in the steady vibration state with respect to the sine or cosine input corresponds to the derating line of M (φ, η), and the sign of equation (31) is +. In addition, when ξ = π, 3π / 2, 2π, φ = −φ sequentially ₀ , 0, φ ₀ Therefore, it corresponds to the force line of M (φ, η), and the sign of the equation (31) is − (see Table 1).
[0052]
[Table 1]

Using this, equation (30) can be modified as follows.
[0053]
[Expression 32]

In the equation (32),
[0054]
[Expression 33]

R (φ, η) is called a sum curve. Further, since the history curve for the steady solution is symmetric with respect to the origin, the equation (32) can be simplified as follows.
[0055]
[Expression 34]

Since P (φ) in the above equations is a constant function, R (φ, η), that is, the sum curve obtained by adding the force line and the force reduction line at the same angle level is the same. If there is, C (φ ₀ , η) are equal. Where A (φ ₀ , η), R (φ ₀ , η) as a basic model of the history curve, the power-type restoring force model shown in FIG. 8 is applied. The basic formula can be expressed by the following formula.
[0056]
[Expression 35]

Where α and k are φ ₀ And a function of η.
[0057]
In the restoring force model shown in FIG. 8, the area A (φ ₀ , η) can be calculated as:
[0058]
[Expression 36]

Based on the bone curve of equation (35) and the equation for calculating the area of equation (36), α and k are the following equations.
[0059]
[Expression 37]

[Formula 38]

Also, when substituting equation (36) into equation (29),
[0060]
[39]

It becomes. The sum curve R (φ, η) is
[0061]
[Formula 40]

Therefore, from equation (34), C (φ ₀ , η) can be obtained as:
[Outside 4]

[Expression 41]

In the equation (41), Γ represents a Γ function.
[0062]
If the equations (39) and (41) are substituted into the equations (26) and (27), respectively, the equivalent spring constant K _eq , Equivalent damping coefficient H _eq Is required as follows.
[0063]
[Expression 42]

[Equation 43]

That is, if α and k of the power function type restoring force model are obtained, the equivalent spring constant K _eq , Equivalent damping coefficient H _eq Will be determined.
[0064]
Next, how to determine the shape parameters α and k of the power functional restoring force model will be described. Here, a method of replacing the basic loop of the history curve with the power function type restoring force model is described.
[0065]
The following two conditions reflect the vibration characteristics.
(I) The area surrounded by the hysteresis curve obtained from the dynamic force test is made equal to the area surrounded by the hysteresis curve of the power function type restoring force model (matching of attenuation).
(II) The bone curve obtained by connecting the vertices of the history curve obtained from the dynamic force test is matched with the bone curve of the power function type restoring force model. That is, the vertex of the dynamic force test and the restoring force model are matched at each displacement level (matching of rigidity).
[0066]
Now, in the hysteresis loop obtained from the dynamic force test, the area surrounded by the hysteresis curve and the bone curve are changed to the steering angle φ. ₀ And the external force frequency η ₀ (Φ ₀ , η) and F ₀ (Φ ₀ , η). However, G ₀ (Φ ₀ , η) and F ₀ (Φ ₀ , η) is θ _s , M _s It is assumed to be dimensionless.
[0067]
The condition (I) can be expressed by the following equation using the equation (34).
[0068]
(44)

The condition (II) can be expressed by the following equation from the basic equation of the bone curve of equation (33).
[0069]
[Equation 45]

If this equation (45) is substituted into equation (44) and k is eliminated, the following equation is obtained.
[0070]
[Equation 46]

From the equation (46), α is obtained as the following equation.
[0071]
[Equation 47]

Moreover, k is calculated | required like following Formula from (45) Formula.
[0072]
[Formula 48]

[0073]
Α (φ expressed by equations (47) and (48) ₀ , η) and k (φ ₀ , η) can be used to match the area and bone curve enclosed by the respective hysteresis curves in the dynamic force test and the power function type restoring force model. That is, G ₀ , F ₀ Steering angle φ ₀ If the function can be expressed by the external force frequency η, α (φ ₀ , η) and k (φ ₀ , η) can be used to replace the power function type restoring force model. This α (φ ₀ , η) and k (φ ₀ , η) is substituted into Eqs. (42) and (43), and the equivalent spring constant K _eq And equivalent damping coefficient H _eq From the following equations (49) and (50), the spring constant K (θ that depends on the steering angle θ and the steering speed (frequency) ω of the steering system ₀ , ω ₀ ) And damping coefficient C (θ ₀ , ω ₀ ).
[Formula 49]

[Equation 50]

Next, it shows about the measurement result confirmed using the actual vehicle. First, using a car A with good steering and a car B, which is considered to be bad, a flat dry asphalt road surface is sine by changing the maximum steering angle and the steering speed (frequency) stepwise at a constant vehicle speed of 140 km / h, for example. Wave slalom running was performed, and a steering angle-steering torque hysteresis curve for each steering frequency was measured. For example, vehicle A is as shown in FIG. 3B, and vehicle B is as shown in FIG. Specifically, the steering angle steering torque hysteresis curve is measured when the steering frequency is every 0.25, 0.5, 0.75, 1.0 Hz and the steering angle is 5, 10, 15, 20 degrees, Hysteresis curve to bone curve M ₀ (Φ, η) and the area S surrounded by the hysteresis curve ₀ (Φ, η) was determined. This bone curve may correspond to the dynamic spring constant and the area may correspond to the damping coefficient.
[0074]
Calculated bone curve M ₀ (Φ, η) and area S ₀ Using (φ, η), the values of α and k were determined from the equations (47) and (48). As an example, FIG. 9 shows the values of α and k with respect to the steering angle when the steering frequency is 0.5 Hz, with the A car indicated by a circle and the B car indicated by a cross. FIG. 9A shows α with respect to the steering angle, and FIG. 9B shows k with respect to the steering angle.
[0075]
Further, the steering spring constant was obtained from the equations (26) and (50) using α and k obtained as described above. FIG. 10 shows a comparison of the steering spring constant A car (circle) and B car (cross mark) with respect to the steering angle with a steering frequency of 0.25 Hz. As shown in the drawing, the steering spring constant K is low in the vehicle B which is poor with respect to the vehicle A with good steering performance in a situation where the steering speed is moderate in a region where the steering angle is small (the situation where the steering frequency is low). I understand that. This quantitatively expresses the lack of response at the time of steering in this region, and can be said to be a major factor that deteriorates the steering performance of the B car.
[0076]
Similarly, the steering damping coefficient was obtained from equations (27) and (51) using α and k. FIG. 11 shows a comparison between a car A (circle) and a car B (cross) with respect to a steering angle with a steering frequency of 0.25 Hz. As shown in the figure, it can be seen that the B car has a smaller steering damping coefficient in all regions than the A car with good steering performance. This expresses the lack of stability in a broad sense including the lack of positioning at the time of steering, and it can be said that this is also a major factor that deteriorates the steerability of the B car. The same applies to the steering frequencies of 0.5, 0.75, and 1.0 Hz.
[0077]
In this way, the power function type restoring force model is used as a method for determining the nonlinear vibration characteristics (spring constant, damping coefficient) of the steering system from the measured value of the steering angle-steering torque hysteresis curve for each steering frequency during actual vehicle travel. Using the equivalent linear system analysis method, the car with good steering performance and the car with poor steering performance were analyzed. As a result, it was found that the nonlinear vibration characteristic of the steering system that greatly depends on the steering frequency and the steering angle is required. In addition, from the results of analysis using this analysis method, it becomes possible to quantitatively grasp the nonlinearity of the steering force characteristic, which was difficult with the conventional method. I was able to express. From the above, a new analysis method could be proposed as a method for quantitatively analyzing the steering performance.
[0078]
In the illustrated example, the steerability vehicle evaluation device is shown. However, the vehicle evaluation device to which the present invention is applied is not limited to the steerability, and can be regarded as a dynamic spring and can be digitized. It can be applied to the entire vehicle such as brake evaluation, suspension evaluation, and acceleration performance evaluation. This is particularly effective for those exhibiting nonlinear characteristics.
[0079]
【The invention's effect】
As described above, according to the present invention, a hysteresis curve is created by the input change value and the input corresponding change value that changes in accordance with the input, and the hysteresis curve is characterized by two dynamic spring constants and a damping coefficient. Since it can be expressed by an index value, sufficient vehicle evaluation can be performed even for a vehicle that exhibits strong nonlinearity. Therefore, even if the steering torque (reaction force transmitted to the hand via the steering) and yaw rate with respect to the steering angle and steering speed when the driver operates the steering and the yaw rate have nonlinear characteristics, the nonlinear characteristics are satisfactory. Can be grasped. This makes it possible to quantitatively grasp even the above-mentioned nonlinear characteristics that could not be analyzed by the conventional evaluation method, and perform vehicle evaluation of steering performance, which is one of the important performances of automobiles. Can now.
[0080]
In addition, by creating a hysteresis curve using any two of steering angle, steering torque, yaw rate, and lateral acceleration as input and the corresponding change values, steering performance can be evaluated as vehicle evaluation. This makes it possible to evaluate the steering performance, which has been difficult, using two index values, and to easily achieve a desired steering feeling.
[0081]
Also, for example, the input amount of the input amount dependent characteristic is the steering angle magnitude (amplitude), the input frequency of the input frequency dependent characteristic is the steering speed (frequency), and the dynamic spring constant and the damping coefficient are calculated from these input values. can do. As a result, the non-linear vibration characteristics (spring constant and damping coefficient) of the steering system depending on the steering angle and the steering speed can be obtained from the restoring force characteristics obtained by conducting the steering force experiment in which the steering speed during actual vehicle running is changed stepwise. The difference in vehicle characteristics can be clearly expressed from the result.
[Brief description of the drawings]
FIG. 1 is a schematic plan view of a vehicle to which the present invention is applied.
FIG. 2 is a schematic diagram of control by a control unit.
FIG. 3 is a hysteresis curve of steering angle-steering torque, where (a) is before adjustment of an adjustment target vehicle, (b) is a target vehicle, and (c) is after adjustment.
FIG. 4 is an evaluation index map of a spring constant K and a damping coefficient C in steering angle-steering torque.
FIGS. 5A and 5B are steering angle-yaw rate hysteresis curves, in which FIG. 5A is before adjustment of an adjustment target vehicle, FIG. 5B is a target vehicle, and FIG. 5C is after adjustment;
FIG. 6 is an evaluation index map of a spring constant K and a damping coefficient C in steering angle-yaw rate.
FIG. 7 is a diagram showing a steering vibration model.
FIG. 8 is a diagram illustrating a power function type restoring force model.
9A is a diagram illustrating α with respect to a steering angle for obtaining a steering spring constant in a vehicle A with good steering performance and a vehicle B with poor steering performance, and FIG. 9B is a diagram illustrating k with respect to the steering angle.
FIG. 10 is a diagram showing a spring constant with respect to a steering angle at a steering frequency of 0.25 Hz in a car A having good steering performance and a car B having poor steering performance.
FIG. 11 is a diagram showing an attenuation coefficient with respect to a steering angle at a steering frequency of 0.25 Hz in a car A having good steering performance and a car B having poor steering performance.
[Explanation of symbols]
7a Lissajous generator (hysteresis curve generator)
7b Constant calculation unit (dynamic spring constant calculation means / damping coefficient calculation means)

Claims (1)

A vehicle evaluation device for determining the assist gain and damping gain of the evaluation to the vehicle steering against any input as a control parameter of the steering system of the vehicle I quantitatively line,
A hysteresis curve creating means for creating a hysteresis curve as the input corresponding change value for changing any two of the steering torque and the yaw rate and lateral acceleration and the steering angle in response to the input and the change value of the input,
Dynamic spring constant calculating means for calculating a dynamic spring constant in vibration characteristics from the slope of the hysteresis curve;
A damping coefficient calculating means for calculating a damping coefficient in vibration characteristics from the area of the hysteresis curve ;
The assist gain and the target spring constant corresponding to the dynamic spring constant of the target vehicle, the target damping coefficient corresponding to the damping coefficient, and the deviation obtained by comparing the dynamic spring constant and the damping coefficient, respectively. A vehicle evaluation device comprising means for obtaining a damping gain .

Images