Reliable and accurate short-term traffic forecasting system is crucial in supporting any Intelligent Transportation System. The past two decades have witnessed many forecasting models being developed, yet none of them could consistently outperform the others under various traffic conditions. To deal with the nonlinearity and non-stationarity of dynamic traffic process, a real time neural network learning approach is taken and a traffic flow mode based forecasting method is presented. Results obtained from case study indicate the proposed approach can enhance adaptability of short-term traffic forecasting and has the advantages of better flexibility and transferability.
展开▼
