Research On The Optimal Layout Of Traffic Detector For OD Trip Matrix In Road Networks

The layout of traffic detector in road network is the key technology of the urban traffic information collection system, also the foundation of the intelligent transportation system. The study of layout optimization of urban road traffic detectors is useful for taking advantage of the limited traffic information collection facilities, in order to reduce the traffic information acquisition cost, improve the capacity of traffic information collection, and provide the basic data for the implementation of effective traffic management.As the basic input data in the traffic simulation system, the OD trip matrix is critical for traffic management and control. However, traditional OD matrix acquisition method is costly. With the development of urban traffic information collection technology, the information of link traffic counts is more easily accessible, so estimating origin-destination trip matrices from link traffic counts has been a subject of substantial research.Previous studies have overwhelmingly focused on the development of the traffic counting point location, while paying very limited attention to the difference of the layout cost in every location point. In this paper, on the basis study of OD matrix estimation model and layout optimization of traffic detector, taking the differences of the layout cost in every location point, the model is established in which the cost of the detector layout is the objective function and the OD information cover is the restriction. Then, the paper analyzes the operation principle, applicable aspect, layout cost of different types of traffic detectors, and established models for different types of traffic detector. Loop detector based on its mature technology and low cost advantages, is widely used in the urban road detection technology. From the layout technology of loop detector, the number of the loop detectors in the link is proportional to lanes number of the link. So if the number of the lane information in the road networks is available, the model is more accurate in the layout cost aspect. To solving the model, this paper use the genetic algorithm and Matlab software programming. Finally, this paper is through a small network to validate the reliability of the model, and through a large network to validate the feasibility of the model.