Research On The Traffic Self-Adaptive Of Subway Operation Dispatch Strategy

The processing of urbanization and the increasing of the city’s population bring new pressure to the city’s rail transit. The passenger flow of rail transit stays very high, which bring new problem to the management of the operation department. When the train runs frequently, a little fault of the system will cause massive delay, which have a big influence on the passengers’ trip. So the prediction of passenger flow and the establishing of operation plan is important issue in the rail transit research. Under the consideration of passenger flow’s dynamic change and the constraints of train and line, this thesis has improved the comfort level of the passengers, decreased the operation cost by using the dispatch strategy reasonably.First of all, the rail transit passenger flow was analyzed by this thesis. Its distribution rule was also studied respectively from the time and space to find out the law of it, which could be helpful for the short-term passenger flow prediction. The train routing plans were classified and their advantages and disadvantages were analyzed in this thesis. The passenger distribution of Changsha second metro line was analyzed from the time and space,which provided a theoretical basis for the passenger flow forecast.Secondly, based on historical data of AFC system, a short-term passenger flow forecast of future was realized by the combination forecast model. Compared with the single prediction model, the combination of model had a higher prediction precision than single prediction model.Finally, a plan for the optimized operation was worked out through setting the size of the train routing, based on the prediction of passenger flow and with the goal of total mileage and full rate target. Compared with the original operating plan, the optimized plan improved the quality of service to passengers and train load factor. At the same time, for the urban rail transit operations, the plan not only reduced the number of on-line train but also accelerated the recycle rate of carriages and saved operating costs.This thesis has made a subway operation plan based on the passenger flow, this plan can make a good prediction to the passenger flow during a future time through the mixed prediction model, then the operation department will make the optimal choice. The total miles of the train has decreased by 14.51%, and the total number of running train has decreased by 11.9% after the optimization,which has saved the operation cost of the rail transit. Besides, the full load rates of different section during peak time have all decreased in different degree, the comfort level of passengers has been improved, which has decreased the crowdedness. Therefore, this thesis has some application value to the rail transit operation department.