Study On Dispatch Strategies Of Subway Train

Subway, as an important constituent of transportation in major cities, has played an irreplaceable role in alleviating road congestion and promoting the development of urban construction. With the gradual increase of traffic and competition, it is difficult to meet the demand of passenger and to adjust the change of traffic in the subway system. The traditional mode of subway operation is simple and the single dispatch will not change for a long time, which will cause variety of problems such as long waiting time of passenger and a high empty-loading ratio of train in at off-peak hours, overcrowding of carriage and shortage of transport capacity in rush hours. In order to improve this situation, the article focuses on the research of subway passenger flow forecasting and dispatch optimization.First of all, the article analyze the source and influence factors of subway passenger flow, getting the distribution characteristics of passenger flow in aspect of time and space. In terms of time, the passenger flow distributions on day, week and seasons are respectively discussed. For space, the passenger flow distributions on line direction and cross section of subway are analyzed. It can be used as the basis for dispatch optimization via quantitative analysis of imbalance of passenger flow.Secondly, a method based on least squares support vector machine flow is chosen to quantitatively predict subway passenger flow. Considering the passenger flow being easily affected by external factors, we transform qualitative factor into the quantitative one which is easy to be calculated, normalize the factors by combining history passenger flow, and put the data into the prediction model to have train and forecast and finally get prediction result by inverse transformation.Thirdly, different types of subway train and marshaling plans are compared and the set standard of train interval is analyzed in time division dispatch optimization. The advantages and disadvantages, set conditions and train turn-back modes of different routing plan are also analyzed in peak time dispatch optimization.Finally, it determines on necessities of dispatch optimization on the basis of temporal and spatial imbalance in predicting passenger flow. The multi-objective optimization model is established under the goal of reducing the waiting time of passengers, keeping comfort and improving load factor. According to different configuration types of train, the article proposes a method based on multi-objective coordinating optimization to solve the model, and then it gets dynamic running interval optimization strategies of variable marshaling and fixed one. The analysis results about passenger flow at peak time show the imbalance is very high. In order to solve the problem, optimized routing is determined to set up optimization mode. After calculation it can gets dynamic optimization strategy of routing plan. In the end, the level of service and operating costs in three kinds of dispatch strategies are compared and the applicability about the strategies for different subway operating companies is analyzed, which can provide theoretical supports and suggestions for executives.