摘要
[目的]为实现列车牵引制动关键系统的可靠控制,同时抑制通信时延对列车控制产生的影响,针对高速列车关键网络系统提出一种离散RBF(径向基函数)自适应鲁棒滑模控制方法。[方法]介绍了高速列车牵引制动网络系统;设计了考虑时延分数项的自适应鲁棒滑模控制方法,并对其进行了稳定性分析;在测试平台上利用组态软件对所提离散RBF自适应鲁棒滑模控制方法进行联合仿真分析。[结果及结论]将离散系统中的时延分为整数项和分数项,在充分考虑时延分数项的基础上,推导出离散趋近律下的时延补偿滑模牵引制动力,其中的未知非线性函数利用具有自适应调节特性的RBF神经网络进行精确逼近。为抑制列车运行过程中受到的较强干扰,将基于干扰观测器的滑模牵引制动力融入模型中,以增强其抗干扰性能。所提离散RBF自适应鲁棒滑模控制方法在稳定性能和响应性能方面优于其他控制方法,具有更为理想的时延补偿效果和鲁棒性能。
[Objective]To achieve reliable control of traction and braking key systems in trains while suppressing the impact of communication delay on train control,a discrete RBF(radial basis function)adaptive robust sliding mode control(abbreviated as ARSMC)method is proposed for critical network systems in high-speed trains.[Method]The traction and braking network systems in high-speed trains are introduced.An ARSMC method considering fractional delay terms is designed,and its stability is analyzed.The proposed discrete RBF ARSMC method is jointly simulated and analyzed using configuration software on a testing platform.[Result&Conclusion]The delay in discrete systems is divided into integer and fractional terms.By fully considering the fractional delay terms,the sliding mode traction and braking force with delay compensation under discrete approximate law is derived.The unknown nonlinear function in which is accurately approximated using an RBF neural network with adaptive adjustment characteristics.To suppress strong disturbances encountered during train operation,the sliding mode traction and braking force based on disturbance observer is integrated into the model to enhance its disturbance rejection performance.The proposed discrete RBF ARSMC method demonstrates superior stability and responsiveness compared to other control methods,exhibiting more ideal delay compensation effect and robust performance.
作者
刘勇
张彤
赵科
LIU Yong;ZHANG Tong;ZHAO Ke(Artificial Intelligence Key Laboratory of Sichuan Province,644000,Yibin,China;College of Locomotive and Rolling Stock Engineering,Dalian Jiaotong University,116028,Dalian,China;College of Automation and Electrical Engineering,Dalian Jiaotong University,116028,Dalian,China)
出处
《城市轨道交通研究》
北大核心
2024年第7期43-48,共6页
Urban Mass Transit
基金
人工智能四川省重点实验室开发基金项目(2021RZJ04)。
关键词
高速列车
滑模网络控制
RBF神经网络
high-speed train
sliding mode network control
RBF neural network
作者简介
第一作者:刘勇,实验师,112554310@qq.com;通信作者:张彤,副教授,zhang_tong66@126.com。
