车辆侧向运动的人工神经网络辨识

Identification of Vehicle Lateral Motion Based on Artificial Neural Network

利用一个带有循环的前馈多层人工神经网络模型对车辆的侧向操纵运动进行辨识。该辨识网络经过训练之后,可以较精确地描述车辆运动系统的输入—输出非线性动态映射关系,从而为车辆运动学建模提供了一条新的有效途径。

The vehicle lateral motion is identified based on a feed-forward multilayered neural network with state recursion. After trained on measured data, the net can accurately approximate the input-output nonlinear dynamic mapping relationship of the vehicle system. So, the neural network identification offers a new effective approach to the vehicle kinematics modelling.