Study On Passenger Flow Forecast And Train Plan Of Urban Rail Transit

At present, the passenger flow forecast of urban rail transit is not accurate during the planning stage, which affects the making of train plan. Studying on the short-time passenger flow forecast and train plan of urban rail transit plays an important role in meeting passenger travel and improving service level.This paper mainly talks about passenger flow forecast, train plan making and the system that can forecast the passenger flow and make the train plan automatically.Firstly, after the short-term passenger flow forecasting method of urban rail transit was analyzed, two representative passenger flow prediction models were established, one was the gray model, which has four old data, and the other was the BP neural network, which has one input node, one output node and three layers.Secondly, the train plan meaning of urban rail transit was summarized. This paper focused on daily operation plan and crossing, the steps to make daily operation plan and crossing were concluded.Based on the above research, the system, which can forecast the passenger flow and make the train plan automatically, was designed. With the help of the Visual Studio 2010 and C#, we developed the system. The system has four major functions, data management, short-term passenger flow forecast, error analysis as well as the plan making based on traffic predictions.Finally, with the Chengdu metro line 1 as an example, we found the development system was available. Compared with gray model, we found that the BP neural network model had higher prediction accuracy. That is to say, the BP neural network is more suitable for the passenger flow forecast of urban rail transit, which has non-linear and uncertainty characteristics.