| With the continuous development of high-speed railways,the speed of high-speed trains is getting faster and faster.The coupling relationship between trains and related equipment such as catenaries,wheels and rails is gradually strengthened.The resistance of high-speed trains continues to increase,of which air resistance The proportion is getting higher and higher,and the aerodynamics is highly nonlinear and uncertain.These factors make the nonlinearity and uncertainty of the trains more and more obvious during the driving process.In order to ensure the safety,punctuality and stability of high-speed trains The operation of high-speed trains with better performance needs to be studied in terms of modeling and control methods.Related researches have been done on the modeling and control of high-speed trains.The thesis has carried out the following research contents:(1)Aiming at the problem of non-linear modeling of high-speed train operation process,a modeling method of high-speed train operation process based on forecast error method is proposed.Taking the high-speed train as the research object,the dynamic system of the high-speed train is analyzed,and the state space model of the high-speed train’s running process is constructed.Based on the idea of system identification,according to the actual input and output data of the high-speed train operation process,the prediction error identification method is used to identify the state-space model of the high-speed train operation process,and the specific identification steps of the prediction error method are given.In order to verify the effectiveness of the algorithm,a simulation experiment was performed with the subspace identification method under various noise conditions to verify the model error.The simulation results show that the prediction error method has a certain anti-interference ability,and the accuracy of the high-speed train system model it establishes High,in line with the actual situation of high-speed train operation process.(2)According to the state-space model of the high-speed train operation process identified by the prediction error method,a state feedback predictive control method based on the prediction error method is proposed.Use the state variable to form the feedback law to solve the predictive control law,use the difference between the actual speed and the predicted speed to feedback correct the speed prediction value at the next moment,design the state feedback prediction controller,and complete the high-speed train target speed curve high-precision tracking.control.In order to verify the effectiveness of the algorithm,simulation comparison experiments were carried out with PID control and generalized predictive control.The simulation results show that the state feedback predictive control method has higher tracking accuracy and better control performance,which meets the safety,punctuality and stability of high-speed trains.Operational requirements.(3)Aiming at the problems of nonlinearity and uncertainty ignored or simplified when establishing the state space model of high-speed trains,a high-speed train BP neural network was designed based on the high-speed train state feedback predictive control method based on the prediction error method.The online compensator compensates the non-linearity and uncertainty of the high-speed train identification model and improves the tracking accuracy of the control system.Use the current state of the system and the actual speed as the input of the BP neural network online compensator,refer to the performance index function formed by the difference between the trajectory and the actual speed,perform online training,and output the compensation control force to the control system to realize the high-speed train Online compensation of control force during operation.Simulation results show that the addition of the BP neural network online compensator can further improve the speed tracking accuracy of high-speed trains,and better achieve the requirements of safe,punctual and stable operation of high-speed trains. |
