基于滑模和径向基函数神经网络的城市轨道交通列车速度跟踪控制算法

Research on Speed Tracking Control Algorithm for Urban Rail Transit Trains Based on Sliding Mode and RBF Neural Network

[目的]针对城市轨道交通列车运行控制系统中传统ATO(列车自动运行)速度控制算法在速度跟踪控制方面存在的控制精度不高和抗扰动性差的问题,提出了一种新的速度控制算法来提高控制精度。[方法]首先,建立了列车运行的单质点动力学方程,针对牵引和制动系统在执行指令时存在的时滞现象设计了时延补偿模块;其次,在控制器设计部分采集速度和位置误差建立滑模切换函数,并通过微分方程推导建立滑模控制器;最后,为了抑制滑模控制器固有的抖振现象,将其输出的切换控制量采用径向基神经网络进行目标训练从而优化控制器。[结果及结论]基于徐州地铁3号线二期改造列车参数在Matlab软件上进行仿真试验,其仿真结果证明该算法保证了列车在运行过程中,控制器输出的速度可以更高效精确地跟踪给定的推荐速度曲线。

[Objective]Addressing the issues of low control accuracy and poor disturbance rejection in conventional ATO(automatic train operation)speed control algorithms in urban rail transit train operation control systems,a new speed control algorithm is proposed to improve control accuracy.[Method]Firstly,a single-mass point dynamic equation for train is established,and a delay compensation module is designed to address the phenomenon of delay in executing commands by the traction and braking systems.Secondly,in the controller design part,speed and position errors are collected to establish a sliding mode switching function,and a sliding mode controller is derived through differential equations.Finally,to suppress the inherent oscillation phenomenon of the sliding mode controller,the switching control output is optimized by training a RBF(radial basis function)neural network.[Result&Conclusion]Simulation experiments are conducted based on the parameters of the train from the Phase II renovation of Xuzhou Metro Line 3 in Matlab software.The simulation results demonstrate that the proposed algorithm ensures that the controller output speed can more efficiently and accurately track the recommended speed curve during train operation.