Bus Passenger Flow Analysis And Prediction Based On Transport IC Card Big Data

With the rapid development of ITS in China, urban transport IC card is widely used thanks to its convenience, rapidity, inter-district and other advantages. The crucial part for intelligent bus system lies on the comprehensive, accurate forecasting of bus passenger flow, thus bus passenger flow analysis and prediction based on the transport IC card big data is of great significance.In this paper, the public transportation passenger flow is analyzed and forecasted which is based on transport IC card big data that through data processing, and some uncertain factors such as line type, date, time, weather condition, temperature and wind are involved. Three kinds of model, the multivariate linear regression model, support vector machine regression(SVR) model and the fuzzy neural network model which are commonly used in bus passenger flow prediction are compared during modeling. The prediction precision of each model are compared after simulation, the results show that the multivariate linear regression model can well realize bus passenger flow forecast based on urban public transport IC card big data. Finally the applicability and accuracy of the multivariate linear regression model is verified using the city public transport IC card data of Guangdong province. Experiments in real data show that our model has the higher prediction accuracy, and can predict passenger flow rapidly and accurately.In this paper, model is established and bus passenger flow prediction is realized based on mass bus IC card data, weather data and bus routes data. Uncertain factors such as line type, date, time, weather condition, temperature and wind are considered.Problem of unbalanced urban bus lines passenger flow distribution, crowded traffic and so on are solved through bus passenger flow prediction. Big data leads the future city wisdom travel.