Research On Maintenance Optimization Decision Technique Of Urban Rail Transit Track Irregularity Based On Markov Decision Process

In recent years,the development of urban rail transit in China is very rapid.With the continuous increase of urban rail transit operating mileage,the safety and comfort of train operation has attracted more and more attention.The track irregularity has an important impact on the safety,comfort and environmental noise of train operation.Therefore,it is important to study the optimization technology of urban rail transit track irregularity maintenance decision-making to scientifically control the cost of track irregularity maintenance and ensure driving safely.By analyzing the research status of the optimization research on the maintenance decision of track irregularity at home and abroad,combined with the demand for maintenance decision of urban rail transit track irregularity,this paper constructs a Markov Decision Process(MDP)-based urban rail transit track irregularity maintenance decision optimization model.The model divides a line into multiple track units at the length of 200 m.Taking into account the uncertainty and heterogeneity of track irregularity degradation,the MDP method is adopted to formulate the optimal track irregularity maintenance decision in a long period.The model takes the minimum cost from the current decision time to the end of the cycle as the objective function.The constraints include maintenance activity constraints,track unit quantity constraints,maintenance budget constraints,state level proportional constraints and state transition constraints.Finally,the mixed integer linear programming algorithm is adopted to solve the Model.In order to verify the model,the author collected the historical track inspection car data,the annual maintenance plan and the facility account data of line 1 by participating in the Beijing Municipal Transportation Committee project "Big Data-Based Track Intensive Management" and the project of Beijing Subway Operation Co.,Ltd." Beijing Subway Track Intelligent Operation and Maintenance Management Information System" undertaken by Beijing Jiaotong University.The parameters required for the model calculation are determined by the data collected above,and the Monte Carlo-simulation Method was used to simulate the track irregularity changes of the upward line and the downward line of Line 1 for 5 years dynamically,whcih achieved the dynamic optimization of the maintenance decision.Finally,the maintenance decision obtained by the simulation is compared with the actual execution maintenance decision.The results show that the maintenance decision optimized by the model can be scientifically arranged for the track units in different spatial positions based on the track irregular current state and the state change rule,which have effectively reduced the maintenance cost while ensuring safety,thus verifying the effectiveness of the model.The author has designed and developed the Beijing Subway Track Irregularity Maintenance Decision Optimization Subsystem as a subsystem of Beijing Subway Track Intelligent Operation and Maintenance Management Information System.The paper designs the function module and database of the the Beijing Subway Track Irregularity Maintenance Decision Optimization Subsystem,and completes the system development work by Microsoft Visual Studio 2015 development tools and Matlab program.Finally,the application effect analysis of the system is carried out by using the data of the track inspection vehicle on the upline of Line 1,which verifies the effectiveness of the system function.