摘要
提出一种基于粒子群优化与径向基(Radical basis function,RBF)神经网络优化算法的商用车横向稳定性优化控制策略,采用上、下双层控制模式,上层控制器以横摆角速度与质心侧偏角为控制目标,依据车辆行驶工况的反馈信息,利用粒子群优化(Particle swarm optimization,PSO)算法对模糊控制器中的比例因子参数实施动态优化,实现对前轮附加转角和横摆力矩的控制。下层控制器采用RBF神经网络优化制动力分配,通过对横摆角速度偏差的自适应学习,结合滑移率控制器实时优化分配左、右前轮的制动器制动力并修正前轮转角。基于搭建的Truck Sim与Matlab/Simulink联合仿真环境,选取典型试验工况进行车辆横向稳定性仿真分析。研究结果表明,与传统的电子稳定控制系统(Electronic stability control,ESC)控制策略相比较,优化控制后车辆的横摆角速度、质心侧偏角以及侧向加速度等动态响应指标均满足控制要求,并且实际行驶轨迹与目标规划路径之间具有良好的跟随性,有效改善了低附着路面行驶条件下商用车的横向稳定性。
A commercial vehicle lateral stability optimization control strategy based on particle swarm optimization and neural network optimization algorithm is proposed, and upper and lower double control mode is designed, yaw rate velocity and vehicle side slip angle are taken by upper controller as the control target, according to the feedback information of vehicle driving condition, scale factor parameters of fuzzy controller is optimized by particle swarm optimization algorithm dynamically, and the front wheel additional steering angle and yawing movement control is finished. The lower controller utilizes RBF neural network algorithm to make vehicle yaw velocity deviation adaptive learning, and optimizes the allocation of left and right front wheel brake force and corrects front wheel steering angle. Vehicle lateral stability simulation analysis is conducted by typical test conditions under TruckSim and Matlab/Simulink co-simulation environment. The experiment results show that compared with traditional electronic stability control（ESC） strategy, after optimization control the dynamic response values of yaw rate velocity, vehicle side slip angle and lateral acceleration can meet the requirement, and perform well in tracking the target path, comprehensive improve the lateral stability of vehicle under low adhesion road driving conditions.
作者
杨炜
魏朗
刘晶郁
YANG Wei WEI Lang LIU Jingyu(School of Automobile, Chang＇an University, Xi＇an 710064)
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2017年第2期115-123,共9页
Journal of Mechanical Engineering
基金
国家自然科学基金(51278062)
中央高校基本科研业务专项资金(310822161013)资助项目
关键词
汽车工程
商用车
横向控制
粒子群优化
径向基神经网络
TruckSim
联合仿真
automobile engineering
commercial vehicle
lateral control
particle swarm optimization（PSO）
radical basis function（RBF） neural network
TruckSim
co-simulation
