Research On Operation Control Of Urban Rail Train Based On Model Predictive Control

With the vigorous development of the national economy,the rural population is gradually pouring into cities,and the population mobility has also increased the urban transportation demand.Urban rail transit has become a momentous public transportation tool to alleviate urban traffic pressure and improve passenger travel efficiency due to its high passenger capacity and strong transportation efficiency.Compared with high-speed trains,urban rail transit has shorter distance between stations.In addition,urban rail transit must also ensure the punctual,safe and stable operation of trains.Therefore,while ensuring the safety and punctuality of trains,how to make trains operate more efficiently has become a hot topic of research.In view of this,this thesis conducted the following studies:Taking the urban subway as the research carrier,this thesis analyzes the stress of train traction,braking force operation resistance and workshop coupling force in order according to the subway train operation environment and line conditions.Analyze the train operation performance indicators and constraints,establish the corresponding multi-particle train operation model.This thesis analyzes the Model Predictive Control method in detail,uses its advantages of fully considering the train operation constraints and multi-objective optimization to design the controller.The train kinematics equation is converted into the space state expression,and the linear subsection method is used to optimize the running resistance,reducing the calculation amount while retaining the nonlinear characteristics.According to various performance indicators and train constraints of train operation,corresponding objective functions are constructed to improve the controller design.On the basis of MPC the input blocking technology is used to optimize the train controller to make it adopt different control strategies under different operating conditions in view of its disadvantages of large amount of calculation and long time consumption in optimization.By reducing the number of free variables in the model predictive control problem,it can optimize the number of iterations,reduce the amount of computation,ensure the tracking of the operation curve.In the simulation verification stage,the section between Songjiazhuang and Xiaocun stations on the Beijing Yizhuang Line was selected as the experimental route.Based on the SIMULINK simulation environment,the train model,controller design,and optimization scheme proposed in this paper were simulated and verified.Various performance parameters were analyzed to validate the feasibility of the train operation controller using the model predictive control method based on input partitioning technology proposed in this thesis.The test shows that the method proposed in this thesis can track the target speed curve under the condition that the train traction and braking force are within the safe range,it can effectively reduce the calculation time of the controller,while the control effect is improved to a certain extent.