Freeway Incient Management System Design And Algorithm Realization

It is estimated that as high as 60% to 70% of the freeway traffic delay in developed countries is attributed to traffic incident. In order to reduce the negative influence caused by this delay, transportation departments in a number of countries have established freeway incident management system, which makes the freeway under the condition of accidents smooth by planning and coordinating means.With the introduction of the concept of Intelligent Transportation Systems (ITS) in the traffic field, the system has been greatly developed and utilized, and at the same time it boosts the development of urban traffic tremendously throughout the world. At present, the construction of freeway increases with the, rapid development of highway in our country. After being built, it is imperative to develop freeway incident management system in order to take advantage of the utility and operation efficiency.In this paper, firstly the concept of freeway incident and its characteristics are expounded, and the impact of freeway incidents on the traffic flow is analyzed. Based on the present research on freeway incident detection algorithms abroad and in our country, this paper adopts artificial neural network and sets up the corresponding BP network incident detection model according to the principle of artificial neural network. Then in view of system engineering the freeway incident management system is established and a set of steps for developing freeway incident management system are built from incident detection, incident management and control to driver real-time information, which provides traffic managers with reasonable structure and good function. According to the actual demands, the system can not only meet the requsirements of freeway operation, improve the utility, ameliorate the traffic congestion and decrease the traffic accidents, but also increase the economic and social benefit. In the end, incident simulation on the basis of Mat-lab is developed and the performance of this BP network incident detection algorithm is evaluated by simulating the process of incident detection.