The Comparison And Empirical Study On Short-Time Forecasting Method Of Road Network

Short-term predicting of traffic flow is the key to intelligent transportation system which is the most important suggestion to relieve traffic jam, reduce the emission of traffic as well as decrease the traffic accident. This paper begins with the review of previous attempts to forecast the short-term traffic flow. In addition, the comparison among the linear system theory, artificial intelligence method and nonlinear system theory is proposed on the basis of the summarization of the short-time forecasting method of road network. Additionally, the empirical results show that these three methods under multi-cross-sections condition were superior to single-cross-section. Meanwhile, these three methods have their own advantages and disadvantages and applicable scope in the aspect of calculation efficiency, parameter selection and data demand.