The Passenger Phase Forecast Of Subway Station Passenger Capacity

With the advantages of large passenger flow,high speed,small departure interval and little pollution,metro is gradually becoming an essential vehicle,and it could solve the urban traffic jams properly.With the continuous development of the city,the metro network is becoming more and more complex.Passenger flow of the subway station has taken place great changes.The time variation is the most important thing change.Therefore,predicting short-term passenger volume of subway station accurately is very important toeffectively organize passenger flow,make reasonable scheduling scheme,improve passenger service quality and optimize the vehicle departure interval.Based on the real passenger flow data of Xi'an metro line no.2,using the applicable short-term traffic flow prediction,and the characteristics of metro station's passenger,paper forecasts the short-term passenger flow of metro station.Main contents are summarized as follows:(1)Throughdetailed analyzing the time and space distribution characteristics of metro station passenger flow,the paper could come to the conclusion which the short-term passenger flow of metro station has the time imbalance,variation and highly nonlinear characteristics.In addition,data sample types are effectively classified through clustering analysis of passenger flow of metro station,which provides reliable data for forecasting of the short-term passenger flow.(2)Analyzing the merits and demerits.Extensive application of the different theories which forecast short-term passenger flow,there are obvious advantages of BP neural network in predicting short-term passenger flow.Besides,the paper says the basic principle,the network structureand the basic steps of the algorithm for the BP neural network.(3)The paper builds the metro station short-term passenger flow predictionmodel that is based on 30 minutes of short-term passenger flow data of Xi'an metro line 2 FengCheng five road station and the BP neural network,and give detailed network design process.Make use of the trained neural network to forecast the short-term passenger flow of metro station.Results show that the model has the reliable prediction precision.