Research On Traffic State Identification And Prediction Of Highway Network

For the past few years, the backbone of regional highway network in our country has already covered most of the city and greatly reduce the stress of transportation in differernt area. Although the highway network has been preliminarily formed, due to the lack of road network traffic state identification and trend analysis, the road network management and operations are still in relatively backward and passive management, which is hard to ensure the efficient operation of the highway network.It is necessary to research and establish the traffic state identification and prediction system that meet the characteristics of the highway network in our country, through in-depth understanding of the operation mode and current situation of our highway system. In order to improve the efficiency of highway network management and operation, need to study the method of highway network traffic state identification and prediction, which make the operation decision reasonable.Based on summarizing the relevant research about traffic state identification and prediction, this paper firstly analyze the traffic characteristics of the highway network, using the method based on fuzzy clustering to identify the traffic state of the basic road, combining with the road network topological features, proposing a evaluation method of highway network connectivity, on this basis, selecting saturation, flow rate and connectivity of highway network as index, using the fuzzy comprehensive evaluation method to achieve highway network traffic state identification; Secondly, by analyzing and mining the highway toll data, using BP neural network and kalman filtering algorithm for traffic demand forecast, i.e Origin-Destination(OD) matrix forecasting, and combining with the traffic simulation technology to dynamicly assign the OD demand to the network, and then using the traffic state identification algorithm proposed in this paper, to realize the traffic state prediction of the highway network; Finally, taking the highway network in northeast of Beijing as example, to validate the effectiveness of the proposed method in this paper.