摘要
地铁列车在频繁的牵引提速过程中,其加速阶段易发生轮对空转现象。为恢复空转后黏着控制性能并提高列车轮轨黏着利用率,文章对轮对空转再黏着控制进行深入研究。首先,通过搭建列车单轴动力学仿真模型,并结合全维状态观测器和极值搜索算法对列车状态进行估计。其次,提出基于POS-RBF神经网络的滑模控制器。最后,通过仿真对列车空转再黏着控制性能进行验证。仿真结果表明,相比固定切换增益滑模控制器,基于PSO-RBF神经网络可实时调节切换增益的滑模控制器具有更快的再黏着恢复速度、更小的稳定后抖振优点。这一改进有效提升地铁列车在空转时的黏着控制性能,可为列车安全、高效运行提供有力的技术保障。
Metro trains are prone to wheel idling during frequent traction acceleration,especially in the acceleration phase.To improve the control performance of adhesion recovery after wheel idling and increase the adhesion utilization of the train wheels and tracks,this article conducts in-depth research on the control of wheel idling and re-adhesion.Firstly,by building a single-axle dynamic simulation model of the train and estimating the train's state through a full-dimensional state observer and extremum search algorithm.Secondly,a sliding mode controller based on the POS-RBF neural network is proposed.Finally,the simulation is used to validate the control performance of wheel idling and re-adhesion of the train.Simulation results show that,compared to a fixed switch gain sliding mode controller,the PSO-RBF neural network-based sliding mode controller with realtime adjustment of the switch gain has the advantages of faster adhesion recovery speed and smaller poststabilization oscillations.This improvement effectively enhances the adhesion control performance of metro trains during wheel idling,providing strong technical support for the safe and efficient operation of trains.
作者
丁健斌
李世刚
刘敏军
张洋
陈玲帆
DING Jianbin;LI Shigang;LIU Minjun;ZHANG Yang;CHEN Lingfan(Hubei University of Technology School of Electrical and Electronic Engineering,Wuhan Hubei 430070,China;Wuhan Metro Operation Co.,Ltd.,Wuhan Hubei 430030,China)
出处
《现代城市轨道交通》
2024年第8期51-57,共7页
Modern Urban Transit
关键词
地铁列车
最优黏着控制
粒子群算法
径向基神经网络
滑模控制
metro trains
optimal adhesion control
particle swarm algorithm
radial basis function neural network
sliding mode control
