Study On Waiting Passenger Flow Prediction And Bus Scheduling For Hub Transfer

This paper mainly focused on the situation that the waiting passenger flow of bus hub stations is large and other stations is relatively small, carried out a prediction about waiting passenger flow. Then the prediction results of passenger flow can be used for guiding bus departure time and making it more scientific and reasonable.Because the different passenger flow constitution of the hub station, passenger flow presents different characteristics. This paper predicted the waiting passenger flow of bus hub station based on the fractal theory and the gray theory. This paper adopted fractal theory to research the station that passenger flow did not presents obviously features of morning and evening peak but only showed random volatility. Because of the strong random volatility, the traditional linear prediction model was hard to get a satisfied result. This paper used fractal theory to analysis, deal and predict the passenger flow data. The result showed that fractal theory can get a satisfied result, when it was applied to predicting the waiting passenger flow of bus hub station that passenger flow did not present obviously features of morning and evening peak.Because the grey prediction model has higher prediction accuracy when it was applied to the system time sequence presents a monotonic trend, so the paper adopted grey prediction model to predict the waiting passenger flow of bus hub station that passenger flow presents obviously features of morning and evening peak. But the research result showed that the application effect of gray forecasting model were not very ideal. This paper use nonlinear fitting method of least squares method, to predict residual value of GM(1,1), then add prediction value and residual value of GM(1,1), get the final prediction value. The experiment proved that the prediction accuracy of the improved grey prediction model has greatly improved.In the exist study of bus scheduling, mainly through the history of passenger flow collection, analysis and proposed departure timetable. But this paper treats the passenger of hub station and the passenger of general station separately. Firstly, predicted the passenger flow, and then combined the Poisson distribution of the arrival passenger flow, weighed and consider balance the total cost of waiting passengers and the bus company operating costs, created optimization model of bus departure interval, and guidance bus departure time by solving the model.