Research On Automatic Identification Method Of Freeway Traffic Incident In IoV Environment

In recent years,with the steadily increase of the amount of automobiles in China and the world,a series of traffic problems occurs,such as traffic congestion more and more serious,property losses and yearly toll of casualties increases year by year.In recent years,as application and extension of IoT(Internet of Things)in the field of ITS(Intelligence Transportation System),the IoV(Internet of Vehicles)is proposed.IoV improves road traffic efficiency and safety by integrating IoT and ITS technology to realize the coordination of vehicles,roads and pedestrians.IoV is widely attended and studied in the whole world,various technical bottlenecks are broken one by one,commercialization of IoV will be realized in the near future.This paper summarizes and analyses the exiting methods for automatic detection of traffic incidents,and come to the conclusion that almost all of these methods need the help of fixed facilities to collect traffic data.Based on the above background,this paper proposes a simple freeway incident automatic detection method based on set theory in IoV environment-SFID(Set theory based Freeway Incident Detection).This method working does not based on the traffic information collected by roadside fixed sensors,the existence of other vehicles can be sensed only by periodic communication between vehicles in IoV environment,and this data is processed by simple set theory to quickly determine whether or not an incident take place.Therefore,this method not only has the characteristic of simple and high real-time,but also avoids the disadvantages of existing automatic incident detection methods that various fixed sensors must be laid along the road to collect traffic information.Function of the IoV can be summarized as the convergence of various heterogeneous networks to realize efficient and reliable communications.Therefore,this paper firstly make a brief introduction about the system architecture,model,the two major standards used in the world-DSRC and C-V2X,and application scenarios of IoV.Then the detailed theoretical derivation of the SFID method is presented,and theoretical analysis is carried out on several special scenarios such as the reference vehicle(a vehicle selected as the research object)is involved in a real incident and traveling on the road segment near the entrances/exits.Tak:ing into account the powerful matrix computing power of MATLAB software,the theoretical simulation and data analysis of these several scenarios are carried out by it.The obtained conclusions are consistent with the theoretical analysis.Finally,take a time-driven micro-simulation modeling tool widely used in traffic science-VISSIM to carry out realistic simulation of above scenarios,this tool can simulate various processes such as car-following,parking,lane changing,acceleration and deceleration based on the real-time traffic status on road.In this paper,Wiedemann99 is selected as the follow-up model,set the parking lot and deceleration zone on the road to simulate the traffic accidents and congestion/deceleration caused by traffic incidents separately,the vehicle data of each simulation second is output to an offline file,and then is calculated and analyzed following the principle of SFID.The conclusions obtained are consistent with the theoretical analysis.