Research On Traffic Information Collection And Accident Identification Based On Internet Of Vehicles

The Effective collection of real-time traffic information provides highly good quality data for traffic guidance, traffic jams, and accidents detection, which lays a solid foundation for urban transport planning, traffic control and management. Current traffic information is mainly obtained through loop detectors, microwave detectors, video detection equipment and other fixed detection. However, it is unable to satisfy the demand of intelligent transportation system for traffic information due to some shortcomings, such as the presence of high costs of maintenance and installation, the limited range of detection, the difficulty to provide complete network traffic information. RFID technology, with the advantages of multi-target rapid identification, anti-jamming capability, ease of deployment, low cost, durability, and ability of carrying the owner and vehicle’s information, is regarded as a new alternative approach for traffic information collection. The unique edge over other approaches is that all vehicles can be connected to the network by Information Transmission between the RFID readers installed along the road and the vehicle electronic tags. In other words, getting the cars’ number and the owner of static information is available. Therefore, from the theoretical and technical perspective, a comprehensive, in-depth study of the RFID-based internet of vehicles has important theoretical significance and practical value.The purposes of this research are mainly about how to use RFID technology to build the internet of vehicles, traffic information collection based on the internet of vehicles and automatic identification of research, the specific contents are as follow:(1)Application RFID technology to ITS, and building the IoV model using RFID technology. Appropriate parameters of the RFID system are determined according to the characteristics of the IoV. This provides a solid theoretical foundation for studying the IoV.(2)It is necessary to analyze the anti-collision problem existed in the application of the internet of vehicle. Two types of tag anti-collision algorithms and corresponding simulation analyses based on binary tree and the Aloha-based algorithm were carried out. Combined with the practical application of the internet of vehicle, we propose a suitable tag anti-collision algorithm. Using MATLAB software to simulate the acquisition process, the effect of anti-collision algorithm was verified, which provides a better algorithm for solving the collision problem in the application of the internet of vehicle. (3) Detection models and methods which are on the basis of the traditional measurable parameters including traffic flow, density and the average speed and the internet of vehicle measurable parameters including travel time, delay time and probability of selection, were formulated. Using VISSIM simulation software, real network environment as well as the process of gathering information for the internet of vehicle is simulated. Then it is safely concluded that the effect of gathering information is obvious.(4) By analyzing the characteristics of the traffic flow parameters before and after accidents and the feature of the data gathered by the internet of vehicle, automatic identification method absorbed in the queuing theory is proposed. Based on simulation experiments, the performance of the proposed algorithm is analyzed. This proposed algorithm has a high efficiency in detecting accidents and performs well. Meanwhile, it can adapt to different traffic conditions.