基于模糊自适应PID算法的列车自动驾驶系统研究  

Research on Automatic Train Operation System Based on Fuzzy Adaptive PID Algorithm

作　　者：周锐霖 雷成健 刘泽 苏辉亮 ZHOU Ruilin;LEI Chengjian;LIU Ze;SU Huiliang(Hunan CRRC Times Signal&Communication Co.,Ltd.,Changsha,Hunan 410005,China)

机构地区：[1]湖南中车时代通信信号有限公司,湖南长沙410005

出　　处：《控制与信息技术》2023年第3期34-38,共5页CONTROL AND INFORMATION TECHNOLOGY

基　　金：湖南省重点研发计划项目(2022GK2018)。

摘　　要：既有列车自动驾驶(automatic train operation,ATO)系统使用的传统PID算法囿于参数固定,在强耦合、高度非线性的实际运行场景中,难以很好地克服非线性干扰问题,控车效果不佳。对此,文章提出一种基于模糊自适应PID算法的列车自动驾驶方法,其可以根据预先制定的模糊规则实时调整比例、积分和微分系数,使PID控制器具有更强的速度跟踪性能,从而改善列车控制效果。以长沙轨道交通4号线线路数据为算例进行半实物仿真,实验结果显示,采用该算法后,列车真实速度与推荐速度的平均均方根误差为18.876 cm/s,小于传统PID控制器的35.200 cm/s;列车站间旅行贴合度为1.69 m/s,小于传统PID控制器的2.25 m/s,这表明文章所提算法能有效优化推荐速度曲线跟踪能力,提升系统运营效率。The traditional PID algorithm used in the currently existing automatic train operation system is limited due to fixed parameters,making it difficult to achieve excellent control effects in actual operation scenes featuring strong coupling and high nonlinearity,which is mainly attributed to difficulties in overcoming nonlinear disturbances.In light of this,this paper proposes an automatic train operation approach that relies on a fuzzy adaptive PID algorithm,which can adjust the PID parameters in real time according to the preset fuzzy rules,thus improving the PID controller's performance in speed tracking and leading to an improved train control effect.Based on the data from Changsha Rail Transit Line 4,the semi-physical simulation results show that the proposed algorithm resulted in the average root-mean-square error(RMSE)of 18.876 cm/s between the actual train speed and the recommended speed,which is less than the 35.200 cm/s of the traditional PID controller and the value of fitting degree is 1.69 m/s between travel speed and the recommended speed,which is less than the 2.25 m/s of the traditional PID controller,suggesting that the proposed algorithm offers a more effective solution for tracking the recommended speed curves and improving system operation efficiency.

关 键 词：列车自动驾驶系统 PID算法 模糊自适应PID算法 模糊规则 推荐速度 

分 类 号：U264.911[机械工程—车辆工程]