Adaptive Adhesion Control Of High Speed Trains Based On RBF Network Approximation And Robust Fault Tolerance

In recent years, along with the rapid development of High-Speed Trains (HSTs) program, China has taken great breakthrough development in railway construction, railway equipment and carrier services. Because of the wide scale of China high-speed railway network, the long span between north and south, the complicated terrain situation, the various climate situation and the increasing speed of the high-speed trains, the trains operation environment becomes more and more complex and the running safety of the high-speed trains draws more social attention. Objectively, the requirement of the train operation safety is increasing. The running safety of the high-speed trains is closely related to the stability of traction system and the performance of adhesion control system. Therefore, the adhesion control is a key subsystem in the process of trains’operation safety. In this paper, a reliable adhesion antiskid control strategy is studied and designed to guarantee the stability of the trains operation, taking the complicated environment in the period of high speed trains running,the unknown time-varying interference and the traction equipment failures caused by the mechanical fatigue into consideration. The main researches are as the following four parts:(1) First and foremost, the paper expounds the adhesion mechanism, and then analyzes the wheel-rail adhesion process from the microscopic view. The paper elaborates the characteristics of wheel-rail adhesion and main influencing factors, and further studies adhesive anti-slip effect on the process of traction control system. At the same time, taking the complex train running environment into account, the paper puts forward a more accurate nonlinear dynamics high-speed trains model to study the adhesion control system. The accuracy of the proposed model is becoming stronger, it fully embodies that various uncertain time-varying disturbance is imposed on the high-speed trains in practical complex running environment.(2) Considering the high-speed trains need to maintain the adhesion point near the peak level, the optimal adhesion control algorithm in the paper adopts the asymptotic tracking control method to improve the utilization rate of the adhesion between wheels and rails. Therefore, a new improved variable step seeking algorithm is proposed to search the expected track target. The improved variable step seeking method does not depend on the complicated running environment, while only related to the train states. Compared with the previous conventional search algorithm, the improved one has obvious advantages in terms of the accuracy and speed.(3) Although the introduction of the interference in the trains’dynamics model makes the adhesive system more accurate, it also makes the design of the adhesion control strategy more difficult. In order to deal with the unknown time-varying disturbance, the paper adopts the RBF network to approximate the unknown disturbance, and designs adaptive adhesion control strategy to raise the utilization ratio of adhesion between rails and wheels. Finally, the paper gives the proof and analysis for the adhesion control strategy in the prospect of stability and effectiveness.(4) Given the nonlinear of the trains’dynamic model, meanwhile, the paper has taken the traction equipment failures condition into consideration and designed an improved robust adaptive fault-tolerant control algorithm to avoid the abnormity of adhesion control strategy which may be caused by the discontinuity of the controller working point. The paper also makes the stability analysis on the proposed control strategy through the Lyapunov stability analysis method. The designed adhesion control strategy does not rely on a large number of system parameters. It not only has stronger robustness to the uncertainty of system parameters and unknown disturbance of the external operating environment, but also has good fault tolerance to the equipment failures caused by the adhesion control system.In order to validate the effectiveness and stability of the proposed adhesion control strategy, the above solutions have carried out the digital simulation in MATLAB environment. Finally, the paper makes a summary to the main researches in the end and gives the suggestions to the further research work.