Traffic Prediction Road Research Of Merit Based On Wavelet Network

With the rapid development of intelligent transportation system, the intelligent transportation of real-time, accurate and efficient control peremptory has become an important part of intelligent transportation system. Traffic flow prediction as the basis of intelligent transportation system can be induced traffic, road congestion warning and provide effective reference data for the selection of the optimal path. Forecasting accuracy directly affect the traffic management and control results.At the same time, as a result of the existence of road congestion phenomenon, choose a suitable driving path, save driving time also appears especially important. Car navigation becomes more and more popular at present, but it does not choose the time-consuming of the shortest path for human intelligence, therefore, provide optimal moving path traffic induction is of great significance.To improve the accuracy of traffic flow prediction, in this paper, the collection of the original traffic flow data analysis and judgment, after before the error feedback algorithm and neural network based on the analysis of the genetic algorithm to determine the structure of wavelet neural network. Then by improved genetic algorithm is initialized weights of neural network, and the optimal individual is obtained by genetic algorithm in decoding for wavelet neural weights and factor, construct wavelet neural network. Finally established using wavelet neural network to predict traffic flow value, and the prediction error were analyzed, and USES the Dijkstra algorithm to determine the optimal path is the shortest, the model to make use of MFC software implementation.