独立驱动电动汽车转向稳定性控制方法研究  

Research on Steering Stability Control for Independent Driven Electric Vehicle

作　　者：郭烈[1] 许林娜 孙大川 GUO Lie;XU Lin-na;SUN Da-chuan(School of Automotive Engineering,Dalian University of Technology,Liaoning Dalian 116024,China)

机构地区：[1]大连理工大学汽车工程学院,辽宁大连116024

出　　处：《机械设计与制造》2021年第9期66-69,74,共5页Machinery Design & Manufacture

基　　金：国家自然科学基金资助(51575079);辽宁省博士科研启动基金资助(20170520194)。

摘　　要：以提高轮毂电机驱动电动汽车转向稳定性为目的,针对传统PID算法扰动抑制能力不足,利用神经网络提高基于PID的横摆力矩和滑移率控制系统的稳定性,并针对神经网络收敛速度慢、易陷入局部最优解的问题,提出利用粒子群算法对控制器参数进行优化并对权值进行改进的神经网络PID方法。以四轮轮毂电机独立驱动电动汽车为研究对象,以跟踪期望的横摆角速度为控制目标,基于Carsim/Simuink联合仿真平台,对建立的四轮独立驱动电动汽车横向运动学模型及提出的控制策略进行不同工况下的对比验证,结果表明提出的控制方法优化了传统PID控制算法,振动频率幅值小、能更好地逼近理想值,可改善车辆转向性能、提高稳定性以避免事故的发生。To improve the steering stability of electric vehicle driven by hub motors,and aiming at the insufficient perturbation suppression ability of traditional PID algorithm,neural network is adopted to improve the stability of the yaw moment and slip rate control system.Besides,the controller parameters and the weights of the neural network PID method are optimized by the particle swarm optimization algorithm,which helps to figure out the problem of slow convergence and easy tendency to fall into local optimal solution.A four-wheel-motor driven electric vehicle is taken as the research object to track the desired yaw rate value.The transverse kinematics model of the vehicle and the proposed control strategy are modeled and verified under different working conditions via Carsim and Matlab/Simulink co-simulation test platform.Results show that the proposed control method can optimize the traditional PID control algorithm.The vibration frequency amplitude is reduced,and the yaw moment value and slip rate value can better approach the ideal value,which greatly improve the steering performance and stability of the vehicle.

关 键 词：独立驱动 电动汽车 神经网络PID 粒子群算法 转向稳定性 

分 类 号：TH16[机械工程—机械制造及自动化] U469.72[机械工程—车辆工程]