Traffic Incident Detection Based On Macroscopic And Microscopic Traffic Flow Parameters Fusion In Internet Of Vehicles

Traffic incident detection is the basis of ensureing the efficient and safe operation of traffic system.The traditional method of traffic incident detection based on sensors is designed from only macroscopic parameters of traffic flow and these methods are based on single parameter.These methods don’t make full use of the difference of microscopic parameters,which can’t fully reflect the impact of traffic incident on traffic flow and restrict the performances of detection.In internet of vehicles,we can be able to get more and more accurate vehicle information through vehicle/road communication technology.How to make full use of the information to study traffic incident detection based on the fusion of multiple characteristic parameters of traffic flow has important academic significance and application value for improving the performance of incident detection.Therefore,aiming at the above problems,based on macroscopic and microscopic parameters of traffic flow in internet of vehicles,considering the discrete characteristics of microscopic parameters,the paper studys a comprehensive model for distinguishing traffic flow stability based on multi parameters fusion.Then,based on mutation theory,the new traffic incident detection method is studied based on the result of traffic flow stability,in order to improve the performance of incident detection,improve detection rate and reduce false positives.The main contents are:Firstly,the discreteness and change regulation of microscopic parameters of traffic flow are analyzed and studied.Starting from the basic parameters of the vehicles in internet of vehicles,considering the discrete characteristics of microscopic parameters,the discreteness of velocity and space headway are described by the variation coefficient of velocity and the variation coefficient of space headway respectively.The change regulation of microscopic and macroscopic parameters under different traffic conditions are obtained.On the basis of this,the method of traffic flow stability identification is studied based on the discreteness of parameters.In internet of vehicles,we take it as a breakthrough that the traditional method of traffic flow stability identification does not consider the discreteness of traffic flow parameters.Then a comprehensive model for distinguishing traffic flow stability is established based on fuzzy theroy.Finally,the VISSIM simulation is carried out to verify the rationality and effectiveness of the proposed method.Finally,the new traffic incident detection method based on the mutation of traffic flow stability is studied.In internet of vehicles,aiming at the shortcomings of the traditional detection algorithm based on the characteristic parameters of a single macroscopic traffic flow,starting from obtaining the analysis of the relationship between traffic flow stability and traffic incidents,the optimal data aggregation time is determined based on the cross validation mean variance model,which lays the foundation for selecting reasonable data for incident detection.Then a mean change point model is established based on mutation theory,using minimum variance method to search the change point.And automatic incident detection algorithm based on the mutation of traffic flow stability is designed.Finally,the rationality and effectiveness of the algorithm is verified by VISSIM simulation.