基于神经网络的方程式赛车操纵逆动力学研究

Exploring Inverse Dynamics for Handling Formula Racing Car with Neural Network

1.
School of Mechanical Engineering,Beijing Institute of Technology,Beijing 100081

摘要: 以建立某方程式赛车方向盘转角与横摆角速度的非线性映射为目的,通过机械系统动力学仿真软件ADAMS建立了某方程式赛车的虚拟样机模型,建模过程中考虑了轮胎、减振器等非线性因素及空气动力特性。在方向盘角阶跃输入工况下,利用径向基函数神经网络建立了方向盘转角与横摆角速度之间的非线性映射关系。输入的识别结果表明:利用径向基网络进行方程式赛车操纵逆动力学的研究方法是可行的,并且识别精度较高。
- 方程式赛车 /
- 非线性映射 /
- 识别 /
- ADAMS /
- 径向基函数网络
Abstract: We aim to build the nonlinear mapping relationship between the angular velocity and the steering angle of a certain formula racing car.Its virtual prototype model was built by using the simulation software ADAMS.taking into consideration the nonlinear factors such as tire and damping and their aerodynamics characteristics.Under the steering angle step input condition,the nonlinear mapping relationship between the angular velocity and the steering angle was built by using the radial basis function(RBF) neural network.The input identification results show that the above method is not only feasible but also highly accurate.
- formula racing car /
- non-linear mapping /
- prototype model /
- inverse dynamics /
- radial basis function(RBF) neural network

[1]	郭孔辉. 汽车操纵动力学[M].长春:吉林科学技术出版社,1991.
[2]	郭孔辉. 预瞄跟随理论与人一车闭环系统大角度操纵运动仿真[J].汽车工程,1992,(01):1-10.
[3]	Guo K,Guan H. Modeling of driver/vehicle directional control system[J].Vehicle System Dynamic s,1993,(03):141-184.
[4]	许峰,陈怀海,鲍明. 机械振动载荷识别研究的现状与未来[J].中国机械工程,2002,(06):526-531.
[5]	傅志方,饶柱石,周海亭. 一种动态载荷的识别方法[J].上海交通大学学报,1997,(03):5-7.
[6]	尹浩,赵又群,徐培民. 非线性告诉汽车操纵逆动力学建模与仿真[J].农业机械学报,2007,(09):24-33.
[7]	尹浩. 汽车操纵逆动力学的建模与仿真[D].南京:南京航空航天大学,2007.
[8]	张德丰. Matlab神经网络应用设计[M].北京:机械工业出版社,2009.
[9]	张培艳. 基于径向基函数网络的MOTOMAN机械手运动学逆解[J].机械科学与技术,2004,(05):523-525.
[10]	秦冬晨. 面向SUV车型操纵稳定性的多体动力学建模与仿真[J].中国机械工程,2007,(17):2126-2129.