| In recent years,the requirements for system security and reliability are gradually increasing with the rapid development of high-speed trains(HSTs).As the core component of HSTs,the automatic train control system will be very serious if it breaks down.Thus,it is of vital importance to design an advanced fault-tolerant control algorithm for train control system.In this paper,the influences of network time-delay,actuator fault and complex environment on train control system are fully considered.By using sliding mode technology,particle swarm optimization algorithm,neural network and fault-tolerant control theory,the fault-tolerant control of trains and the cooperative control of multiple trains are studied as follows:(1)Based on the self-organizing neural network,a non-singular fast terminal sliding mode control method is proposed for network control of the train key system.According to the successful rate of swarm iteration,an improved particle swarm optimization(IPSO)algorithm containing adaptive parameters and negative gradient position update strategy is designed to optimize the structure and parameters of the radial basis function(RBF)neural network to establish a self-organizing neural network model,which better realizes the nonlinear approximation of the train traction braking process and the accurate prediction of the network delay.Meanwhile,the IPSO algorithm is used to quickly obtain the tuning parameters of controller to suppress the chattering phenomenon from sliding mode control,which ensures that the appropriate control quantity is sent out in advance so that the precise control and delay compensation of the train nonlinear network system are realized.(2)In strong wind environment,an adaptive non-singular fast terminal sliding mode fault-tolerant control method based on neural network observer is proposed for HSTs with actuator faults.The RBF neural network is used to approximate the nonlinear strong wind disturbance,and combined with the error compensation mechanism,the neural network observer was designed to estimate the failure factors of the actuator.Furthermore,the adaptive non-singular fast terminal sliding mode control method with fault tolerance function is obtained by combining the estimated results with the controller.Under different wind conditions,not only the speed tracking control of the nonlinear time-varying system such as HSTs is realized,but also the train has a good fault-tolerant ability to the unknown actuator faults.(3)Based on the actor-critic neural network,a distributed cooperative sliding mode fault-tolerant control method is proposed for multiple high-speed trains(MHSTs)with actuator faults under moving block condition.An adaptive compensation control law is designed to eliminate the influence of unknown actuator faults on the train control system,and the actor-critic neural network is used to estimate the switching gain of distributed sliding mode fault-tolerant controller online to reduce the damage caused by system chattering,which achieves the coordinated operation of MHSTs. |
