四轮转向车辆的径向基函数神经网络复合控制器设计  

Design of radial basis function neural network compound controller for four wheel steering vehicle

作　　者：高琳琳[1] 金立生[1] 郑义[1] 李科勇[1] 

机构地区：[1]吉林大学交通学院,长春130022

出　　处：《吉林大学学报（工学版）》2016年第2期366-372,共7页Journal of Jilin University:Engineering and Technology Edition

基　　金：吉林省科技厅国际合作项目(20130413056GH);长春市科技局重大科技攻关计划项目(长科技合(2013021)号-13KG05);清华大学汽车安全与节能国家重点实验室开放基金项目(KF14182);吉林大学国家杰出青年基金后备人选培育计划项目

摘　　要：建立了四轮转向车辆的八自由度动力学模型,并设计了四轮转向车辆的径向基函数神经网络控制器,该控制器采用直接闭环训练法以离线的方式进行训练。为了克服离线训练导致控制系统适应性差的缺点,本文同时设计了一个单神经元自适应PSD控制器,它与径向基函数神经网络控制器组成复合控制系统共同完成对四轮转向车辆的控制。结合建立的八自由度车辆动力学模型对设计的复合控制器进行仿真实验分析,结果表明,无论匀速工况还是变速工况,径向基函数神经网络复合控制器均可有效地改善四轮转向车辆的操作稳定性。The dynamic model of eight Degrees of Freedom（8D）Four-wheel Steering（4WS）vehicle was established.Using the theory of artificial neural network,a Radial Basis Function Neural Network（RBFNN）controller was designed and it was trained by the direct close-loop training method off-line.In order to overcome the shortcomings of poor adaptability caused by off-line training,a single neuron adaptive PSD controller was also designed.The RBFNN controller and the single neuron adaptive PSD controller composed a compound control system to jointly complete the steering control of 4WS vehicle.Simulation experiments were carried out by combining the RBFNN compound controller with the 8D 4WS vehicle dynamic model.Results show that the designed RBFNN compound controller can effectively improve the steering stability of 4WS vehicle both in constant speed and variable speed conditions.

关 键 词：车辆工程 四轮转向 稳定性控制 径向基函数神经网络 单神经元自适应PSD控制 

分 类 号：U461.91[机械工程—车辆工程]