Generalized Predictive Control Method For Optimization Of High Speed Train

The controller parameters of the generalized predictive control method of traditional high speed train are set according to human experience,and the control effect is affected.In view of this problem,this paper combines the different research methods,and describes the high-speed train as a single power component and a number of power units.Simulated speed tracking control of high-speed trains is carried out by using single input single output(SISO)and multi input multiple output(MIMO)generalized predictive control.And the convex optimization method is combined to determine the value of the parameters of the weighting matrix in the performance index function.The stability of the closed-loop system is taken into account when the tuning parameters are determined.The simulation results show that,compared with the traditional generalized predictive control method,the high speed train has a good tracking ability at a given speed.The detailed research of this thesis is as follows:1,According to the driving and operation characteristics of high-speed EMU in China,it is divided into centralized,single point modeling and distributed multi particle modeling methods,and its advantages and disadvantages are compared.The purpose of this paper is to reduce the complexity of the tuning process of the control parameters.the generalized prediction based on the input traction / braking force and speed data based on the current and past moments Tuning control method for high-speed train power unit known given speed control tracking.By setting up the closed-loop transfer function model of train discrete system,setting appropriate range of prediction step and not long-term control,using YALMIP software to solve the optimization problem to get the relevant tuning parameters,the optimal control input is calculated by minimizing the optimal control input,Combined with the current and past time of the train input traction / braking force and speed data,and will each online calculation of the first control input vector applied to high-speed train system,in each time step after the continuous repetition of the above process,This method achieves better robust closed-loop control performance by tuning the parameters and can accurately track the target curve of high-speed trains.2.In view of the distributed nature of the high-speed trains running on high quality,this paper describes the high-speed trains as multiple power units.Multi-variable input multivariable output generalized predictive tuning control method for high-speed train speed tracking control.Firstly,according to the input traction / braking force and speed data of the current and past time,combined with the discrete train system model and instantaneous input and output data obtained from the identification method,the first step optimal control input is optimized as the objective transfer function,Solve the relevant optimization problems in the paper and find out the tuning parameters,and then get the expected transfer function of the train power unit.Apply the first component of the output speed vector and the first component of the control force vector to the train at each sampling time,This method can achieve better robust closed-loop control performance through continuous rolling optimization.In order to achieve high-speed trains accurately track the target curve.