Passenger Flow Of Public Transport Hub Analysis And Short-term Forecast Based On IC Card

The analysis and prediction of passenger flow in public transport hub provide data support for improving passenger flow organization and security level.Only by making a comprehensive analysis and high precision prediction of the public transport hub,we can master the law of future change of passenger flow and the trend of short-term change of passenger flow in public transport hub,take countermeasures to improve the operation service level and safety level of public transport hub fundamentally,improve the operational efficiency of urban public transport system,and also provide the basis for passenger flow organization and safety early warning scheme in the public transport hub.The main research work of the paper is as follows:(1)The paper determine the composition of passenger flow of public transport hub.Then using data analysis technology analyze the data of bus IC card,to determine the passenger flow,drop-off passenger flow and transfer passenger flow of urban public transport hub swipe card.And by combine with IC card using rate data,it can determine the number of passenger floe,drop-off passenger flow,transfer passenger flow and total passenger floe in public transport hub.(2)The paper analyzes the time variation rule of passenger flow on day,hour and peak hour,and quantitatively analyzes the daily change,hour inhomogeneity and peak hour coefficient of passenger flow in public transport hub,and extracts the short-term characteristics of passenger flow in public transport hub.It provides the forecast thought for the short-time forecast of the passenger flow of the public transport hub.(3)In this paper,BP neural network model and wavelet neural network model are used to predict and evaluate passenger flow of public transport hub in short time.Based on the analysis of the short-term characteristics of the public transport hub,the passenger flow data of the adjacent time period,the adjacent day and the same time period and the adjacent week and the same time period are determined as the input of the prediction model.At the same time,the number of neurons in the neural network models are determined,and the two neural network models are trained and forecasted using the passenger flow data of the public transport hub,and the results are evaluated and analyzed.The prediction accuracy of wavelet neural network models is higher than that of BP neural network model.