高速动车组制动过程的快速广义预测控制

传统高速列车制动方法普遍采用手动控制,实时性不佳,影响控制效果。针对高速列车基础制动装置的结构和动力学特点,建立考虑参数时变、时滞和非线性特征的动车组制动单元模型。考虑高速列车由多个制动单元组成以及各制动单元的相互耦合作用,建立动车组制动过程的多动力单元模型。针对普通广义预测控制计算时间长,实时性差的特点,设计基于极限学习机(ELM)的多输入多输出(MIMO)广义预测控制快速算法对动车组的制动过程进行速度跟踪控制。实现动车组制动过程的准时、舒适、安全和停靠准确。仿真结果表明,相比于普通广义预测控制方法,基于ELM的广义预测快速控制算法使高速列车在制动工况下具有良好的高精度速度跟踪能力,同时具有良好的实时性。

A Fast Generalized Predictive Control Algorithm for Controlling Braking Process of High Speed EMU

The traditional high-speed train braking method generally adopts manual control,and its real-time performance is not good,which affects the control effect.In view of the structural and dynamic characteristics of the high-speed train foundation brake device,the model of the braking unit of EMU was set up,considering the time-varying,time-delay and nonlinear characteristics of the train.Taking account of the high speed train consisting of multiple braking units and the coupling effect of each braking unit,a multi power unit model of the EMU braking process was established.In view of the long time and poor real-time performance of MIMO general generalized predictive control(GPC),a fast algorithm based on the generalized predictive control(ELM)was designed for the speed tracking control of the braking process of EMU.The time,comfort,safety and parking accuracy of the EMU braking process were realized.From the simulation results,we can see that,compared with the general generalized predictive control method,the ELM-based generalized predictive fast control algorithm makes the high-speed train have good high precision speed tracking ability under braking conditions and good real-time performance.