Research On High-speed Railway Train Tracking Control Based On Adaptive Neural Network

Our country has gradually become a high-speed rail power,and the total number of operating lines has reached the first place in the world.The intelligentization of high-speed railways is a development trend.In order to meet the requirements of high safety,high passing capacity and high running performance index of the following vehicles in the mobile blocking mode.In this paper,the problems of uncertain train operation model,limited train operation status,and train measurement signal delay are studied.The specific research contents are as follows:(1)Aiming at the problem of high-speed train tracking control with uncertain train motion models,a robust control method based on a single-parameter adaptive radial basis function(RBF)neural network is designed.The sliding mode method is used to give the ideal control law of train tracking,and the direct robust adaptive RBF neural network is designed to estimate the ideal control law for the uncertainty of the train model.Considering the complexity of RBF neural network weight adjustment,on the basis of adaptive RBF neural network control,the F norm of the upper bound of the RBF neural network weight is taken as the adaptive parameter,and the parameter estimation adaptive law replaces the neural network weight adjustment,Realize single parameter adaptive learning of RBF neural network.The Lyapunov stability analysis ensures that the tracking error of the closed-loop system and the estimation error are consistent and ultimately bounded.Simulation analysis proves the effectiveness of the algorithm.The method is simple in design and convenient for engineering application.(2)Aiming at the problem of restricted train motion model state,a Backstepping train tracking control method based on adaptive RBF neural network is designed.The problems of speed protection and inter-vehicle distance protection during high-speed train operation are discussed,the full-state restricted model of train motion is established,the Backstepping control method of train tracking is given based on the integral obstacle Lyapunov function,and the adaptive RBF neural network estimation virtual is designed.Uncertain function of control law and tracking control law.The Lyapunov stability analysis ensures that the tracking error of the closed-loop system and the estimation error are consistent and ultimately bounded.Simulation analysis proves the effectiveness of the algorithm.This method limits the distance between trains and at the same time realizes the limitation of train speed,which provides theoretical support for the implementation of the mobile blocking method.(3)Aiming at the problem of train signal delay,a train adaptive RBF sliding mode tracking control method with time-varying delay signal estimation is designed.Analyze the characteristics of train time delay and train nonlinear kinematics model,establish a train motion time delay model that satisfies the Lipschitz condition,and design an observer for the train motion time-varying time delay signal.The sliding mode method is adopted to design the train tracking controller,and the uncertain term of the model is estimated by the adaptive RBF neural network.Lyapunov stability analysis guarantees the boundedness of tracking error and neural network estimation error.Experimental simulation verifies the effectiveness of time-varying delay observer and train tracking control.