摘要
速度跟踪控制是列车自动驾驶的一个重要研究方向,但现有速度跟踪控制的性能尚需改进.本文探究了径向基函数(RBF)神经网络PID控制对高速列车速度的跟踪控制.首先改进了高速列车多质点模型;然后研究了被控对象RBF网络整定的PID控制算法,结合反馈控制设计了速度跟踪控制器,根据速度误差用神经网络PID控制决定牵引力和制动力;最后进行了数字化仿真,在仿真过程中考虑了阻力和车厢之间的作用力,并对3种方法进行了对比,结果表明,RBF神经网络PID控制具有最小的速度和距离跟踪误差.相对于单质点模型,本文的多质点模型提高了建模精度,控制系统具有响应速度快、跟踪性能好等优势.
Speed tracking control is an important research direction in the field of automatic train operation(ATO),but the performance of existing speed tracking control needs to be improved.A method of speed tracking control for high-speed trains is studied in this paper based on radial basis function(RBF)neural network PID control.First,a multi-particle model for the high-speed train is improved.Then,the PID control algorithm for RBF network tuning of the controlled object is studied,and the speed tracking controller is designed based on the feedback control.Moreover,the traction force and braking force are determined by the neural network PID control based on the speed error.Finally,a numerical simulation is carried out by using a computer.In the process of simulation,the resistance force and in-train force are also considered,and the three methods are compared by simulation.Simulation results prove that the neural network PID control has the smallest speed tracking error and displacement tracking error.Compared to a single-mass model,the multi-particle model in this paper improves the modeling accuracy.At the same time,the control system has such advantages as fast response speed and excellent tracking performance.
作者
肖龙
梁新荣
王雪奇
董超俊
XIAO Long;LIANG Xin-rong;WANG Xue-qi;DONG Chao-jun(Faculty of Intelligent Manufacturing,Wuyi University,Jiangmen 529020,China)
出处
《五邑大学学报（自然科学版）》
CAS
2020年第2期24-31,共8页
Journal of Wuyi University(Natural Science Edition)
基金
广东省科技计划资助项目(2017A010101019)。
关键词
高速列车
动态模型
速度跟踪
RBF神经网络
反馈控制
High-speed trains
Dynamic models
Speed tracking
RBF neural network
Feedback control
