铁道科学与工程学报2026，Vol.23Issue(2)：542-550,9.DOI:10.19713/j.cnki.43-1423/u.T20250597
基于RBFNN的高速列车分数阶滑模速度跟踪控制
Fractional-order sliding mode speed tracking control of high-speed train based on RBFNN
摘要
Abstract
A speed tracking control strategy,which combines an adaptive RBF neural network and fractional-order sliding mode control,was investigated to address the tracking control problem of high-speed trains subjected to unknown model parameters,resistance uncertainty,and external disturbances.On the basis of constructing a high-speed train dynamics model,a fractional-order nonsingular terminal sliding surface was introduced,and an improved approaching law consisting of both exponential and power convergence terms was designed.Subsequently,a speed tracking strategy for high-speed trains based on the enhanced approaching law was proposed,which can ensure the system status reaches the sliding mode surface within a finite time,improve the system's convergence speed,and suppress chattering.Further,an adaptive control algorithm was used to perform online estimation of unknown parameters such as the basic resistance coefficient and train mass,and the RBF neural network was employed to estimate and compensate for additional resistance and external disturbances,based on which the train speed tracking control strategy was proposed.This enhanced the adaptability and robustness of the train facing time-varying parameter uncertainties,changes in track conditions,and external disturbances.The system's stability was proven based on the Lyapunov stability theory.Simulation verification was performed using CRH380A train parameters.The simulation results show that the speed and displacement tracking errors are minor,convergence is fast,and rapid and precise tracking of the desired speed and displacement is achieved.Compared to the speed-tracking control strategies based on traditional linear sliding mode and integer-order nonsingular terminal sliding mode,the proposed control strategy improves tracking accuracy and enhances convergence speed.关键词
列车速度跟踪控制/分数阶非奇异终端滑模控制/改进趋近律/RBF神经网络/自适应控制Key words
train speed tracking control/fractional order non-singular terminal sliding mode control/improved reaching law/RBF neural network/adaptive control分类
交通工程引用本文复制引用
韩兆玉,徐传芳,高晨旺..基于RBFNN的高速列车分数阶滑模速度跟踪控制[J].铁道科学与工程学报,2026,23(2):542-550,9.基金项目
辽宁省交通科技项目(202318,202320,202344) （202318,202320,202344）
辽宁省教育厅科学研究项目(LJ212510150031,YTMS20230038,JYTMS20230008) （LJ212510150031,YTMS20230038,JYTMS20230008）
