Research On The Prediction Of Intelligent Bus Arrival Time

With the development of the national economy and the automobile holdings increasing, there is huge press on the traffic of cities, and the problems of traffic have become an important factor to restrict the development of the economy of cities. To solve traffic problems in urban should be based on giving priority on public transport system, and the realization of intelligent public transportation system is based on bus information systems, at the same time bus information systems takes "provide quality service for passengers" as the ultimate goal. Therefore required in accordance with the passengers, selective development of public transport system is to enhance and improve the attractiveness of bus travel, and to ease traffic pressure as a effective means. According to a survey revealed that the arrival time for bus is one of the most concerned information to passengers. Therefore, for predicting bus arrival time, the paper analyzes the prediction method, in order to forecast more accurately.This essay starts with the technical and practical feasibility analysis of the bus arrival time prediction technique. After analyzed the existing prediction algorithms' deficiencies, another algorithm will be proposed in this paper, in accordance with the difference between buses routing rime and servicing time, therefore a decentralized state forecast method will be used. In this paper, a large number of researches based on public transport will be done, the factors that affect the prediction results will be analyzed.The paper uses a model, which is based on SVM and re-corrected by real-time speed, to predict the bus routing time; and a algorithm, which is based on parking delayed and recorrcted by exponential smoothing, to predict the servicing time. In this paper, the last, we'll select a T8 bus in Beijing as a case study which uses the model and algorithm designed by this paper. We'll analyze the prediction results and the actual data in order to verify the accuracy of forecasting methods, and reliability.