MTR Short-term Traffic Forecast And Early Warning

With the developing of technology and the expanding of the city size, the MTR as the city's main transport is favored by the majority of people, with its superior performance of punctuality, strong resilience and so on. However, subway station has a limited operating environment, it limited the passenger capacity of MTR. So the excessive influx of passengers inside the subway station, not only will extend the waiting time of passengers, but also may pose a threat to passengers'security in the subway station.The thesis introduced the time series analysis and short-term forecasting of passenger flow. The SAS software is taken to determine the prediction models and the forecast of the passenger flow, and compared the predicted and measured results to determine the forecast error range. And finally the feasibility of short-term forecast is confirmed. Non-transfer subway station in Beijing was taken as the main object of study in this thesis. Through the site investigation and the collation of relevant information, established a model for calculating the short-term passenger flow early-warning threshold. Beijing Zoo subway station was taken as the empirical analysis of short-term forecast and early-warning system, and finally it verified the system's feasibility and practicability.