改进单神经元网络PID算法下的车用轮毂电机控制系统仿真  

Simulation of the vehicle in-wheel motor control system based on the improved single neuron network PID algorithm

作　　者：陈哲明[1] 陶军 庄威洋 钟诚 CHEN Zheming;TAO Jun;ZHUANG Weiyang;ZHONG Cheng(Ministry of Education Key Laboratory of Advanced Manufacturing Technology for Auto Parts,Chongqing University of Technology,Chongqing 400054,China;College of Vehicle Engineering,Chongqing University of Technology,Chongqing 400054,China)

机构地区：[1]重庆理工大学汽车零部件先进制造技术教育部重点实验室,重庆400054 [2]重庆理工大学车辆工程学院,重庆400054

出　　处：《重庆理工大学学报（自然科学）》2022年第5期13-19,共7页Journal of Chongqing University of Technology：Natural Science

基　　金：重庆市自然科学基金项目(cstc2020jcyj-msxmX0226)。

摘　　要：针对采用传统PID控制下轮毂电机控制系统中存在转速超调量大、转矩突变严重等问题,在单神经元网络PID的基础上,引入最优控制理论中的误差二次型性能指标对单神经元网络PID算法的学习规则进行改进,将改进后单神经元网络PID算法运用到轮毂电机矢量控制系统的速度环中。仿真结果表明:改进后单神经元网络PID控制的轮毂电机矢量控制系统速度环响应速度快、转矩突变小以及运行状态更加稳定。The field of new energy vehicles is developing in a prosperous situation.As the tower of strength of new energy vehicles,electric vehicles have received widespread attention at home and abroad.As one of the important components of electric vehicles,wheel hub motors will have a greatly effect on the driver’s driving performance and riding experience.It is well-known that wheel hub motor is a typical multiple variable and highly coupled nonlinear system.Traditional proportional,integral and differential control can not fulfil an ideal control effect for wheel hub motor control system.Aiming at the problems of large speed percent overshoot and large torque fluctuation in the wheel hub motor control system under traditional PID control,in order to solve the drawbacks of wheel hub motor under traditional PID control,the excellent neural network algorithm is considered on the basis of the traditional PID algorithm.The self-learning and self-adaptive performance of the wheel hub motor control system is designed with a PID algorithm based on a single neuron network.The essence of single neuron network PID control is to use proportional coefficient,integral coefficient and differential coefficient as the weight factors of neural network,and to adjust the weight factors through different learning rules,so as to realize the functions of self-adaptation and self-adjustment of the controller.The learning rules have a great influence on the control result of the single neuron network PID.In order to move forward a single step in the control performance of the single neuron network PID algorithm in the wheel hub motor control system,the error quadratic performance index in the optimal control theory is introduced into the learning rule of the single neuron network to carry out the learning rules of the single neuron network PID algorithm.The improved single neuron network PID algorithm is applied to the speed loop of the wheel hub motor field oriented control system,the simulation model of the wheel hub motor and control algo

关 键 词：轮毂电机 单神经元网络 PID控制 误差二次型性能指标 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]