Prediction And Analysis On Short-term Passenger Flow Of Shanghai Metro

Today,with the rapid development of city metro,the passenger flow is increasing at a high speed and is often excessive saturated on some metro lines.Because of lacking the detailed distribution data and the accurate prediction method of passenger flow,it's relatively difficult for the organization of train and the allocation of passenger transport and the early warning of emergency and the establishment of emergency evacuation plan.Therefore,the research on short-term passenger flow forecast and analysis is more and more important.In this article,I studied on the domestic and foreign mainstream short-term passenger flow prediction method in the field of metro and I analyzed the features of time series in the light of historical passenger flow data on workday and holiday of Shanghai Metro.Based on the conclusion of analysis,I established BP neural network prediction model by the date type classification.Then I did some simulation prediction and prediction error analysis with the actual operational data of Shanghai Metro using this model and BP algorithm to verify the effectiveness of the model and the algorithm.Furthermore,combining with the ratio of different path and the traveling time between one station and the next station in the path and the walking time spent on the transfer tunnel,I estimated the passing passenger flow and the flow's destination of one single station.I hope this may provide data support for scheduling other evacuating passengers transportation in the case of sudden interruption of single-station.