Based On Information Fusion Technology Beltway Traffic Flow Guidance System Model

With the development of our country's basic construction and the acceleration of the urbanization, as well as the improvement of the people's living standards and the expansion of private cars, the load of city's transportation increase rapidly. It causes the traffic facilities cannot afford the development of the economic. It is a common problem of the city traffic congestion. It is already proved by the domestic and foreign research that establishing the intelligent transportation system (ITS) is the best method to solve the traffic problem. As a key part of the ITS and the theoretical basis of ITS, the scientific and intelligent traffic guidance system (TGS) plays an important role in solving drivers'reckless driving and in balancing the distribution of traffic flow in the road network.By studying and researching the present situation of the traffic flow guidance system in domestic and foreign, aiming at the expressway of the centre city, the basic structure framework and the specific implementation framework of urban traffic flow guidance system are proposed in this thesis based on the basic architecture of distributed traffic guidance system. And the traffic guidance model was established based on data fusion technology. The predictions of the traffic flow and travel time which were the key parameters of traffic, as well as the dynamic route guidance are researched separately based on the traffic guidance model. For improving the accuracy of guidance, all the basic traffic collection data was fused based on data fusion before forecasting. A dynamic index smooth model based on optimal gradient was used to predict the traffic flow. The real-time performance and the accuracy of prediction were improved based on this model. While on the prediction of travel time, a model of the travel time prediction is proposed which based on the data fusion of the dynamic real-time collection data and the historical traffic data. The traffic information was divided into two parts to deal with separately by this model. On one hand, according to the real-time collection data and the correlation of the traffic parameters, one prediction value of travel time is calculated through the field collection traffic parameters which were fused beforehand; on the other hand, several other values are predicted from the historical data by establishing some prediction models, such as kalman filtering model, gray model, and etc. The final travel time prediction value is fused by the former two types of data. The accuracy of prediction and real-time property is improved revealed by the simulation result and which meets the demand of the traffic guidance. In the traffic flow dynamic route guidance research aspect, the traditional optimal path algorithm was improved in this thesis. According to the current situation of the expressway and take the travel time, travel distance, road occupancy considered as the comprehensive road impedance, an algorithm which fit for the actual traffic flow situation is proposed in this thesis. Finally, a software system is developed for simulation based on the former research. The general functions such as forecasting and publishing the traffic guidance information and optimal route guidance are realized better in this software system.