Study On Real-Time Train Operation Adjustment Algorithm In Urban Rail Transit

At present, China’s urban rail transit construction is in an unprecedented peak. It is understood that there is more than10million people Transported by the Beijing subway every day.As the world’s busiest the train interval of the Beijing subway’s in the peak is close to2minutes. Train interval is relatively small, on one hand, the delay of a train is more likely to cause other trains delay;on the other hand, due to the constraints of the front and rear train, it is more difficult for the delayed trains to return to punctuality. Therefore, the study of real-time and efficient train operation adjustment program has practical significance.The train operation adjustment is worked on the Automatic Train Supervision (ATS) system. The paper is based on the research of ATS system during the author’s internship. This paper discusses the prediction method based on conflict, and forecast the train arrival and departure times as the basis for the adjustment. On the implementation of the adjustment, the unit of adjustment is one train plan, In order to make the trains plan ahead of time and delayed,this paper designs three adjustment modules called the adjustment module of Delaying point,Catching point and Backward offset. In order to make the results satisfy the constraint conditions,the method of detecting and repairing conflict is designed. Then this paper explores the rolling optimization strategy,and meet the requirement of real-time by adjusting in a certain amount of time. Systematic study is made in four aspects of prediction unit,interval running time adjustment unit,Stopping time adjustment unit and unit of conflict detected and repaired.Based on the operational data of Beijing subway line82nd phase, simulation experiment in Different situation is carried out to test and verify the adjust algorithm. It turns out that the adjustment algorithm of this paper can make the trains recovery punctuality quickly when the train is delayed. And it meets the requirements of real time, which proves the feasibility and effectiveness of the adjustment algorithm in this paper.