Study On Freeway Abnormal State Identification Based On The Fusion Of Detector And Toll Data

How the freeway management departments release the traffic information and doroute guidance is all based on the identification of freeway traffic conditions, which isimportant to reduce casualties and life property loss, and to avoid secondary accident.With the development of intelligent transportation technology and the increase of trafficinformation, how to use the detection method of traffic information to automaticallyidentify abnormal traffic state is widely noted. It’s theoretically and practicallysignificant for freeway management departments to improve management level andservice level by studying on the key technology of abnormal state information.In order to solve the problem of ineffective identification of abnormal traffic statedue to the shortage of detector for freeway, the paper establishes the identification offreeway traffic state method, which is based on information fusion and combined withdata source of freeway traffic information collection system. Firstly, the paperselectively analyzes sensitivity of parameter based on detector data, which is under theabnormal traffic state, and establishes the identification of freeway traffic state method,which is based on different data source. At last, by the use of information fusion, thepaper establishes the identification of freeway traffic state method based on fusion ofdetector and toll data, and solves the problem of ineffective identification of abnormaltraffic state due to single data source. The main achievements are as follows:①Sensitivity analysis of parameter. Sensitivity analysis of parameter mainlyconsiders two factors: traffic flow and traffic accident severity, which is under theabnormal state. Sensitivity analysis of parameter laid a solid foundation for the choiceof ACI algorithm; the simulation shows that the traffic state parameter can be sensitiveto reflect the change of traffic state.②To solve the problem of the shortage of detector, the paper analyzes the tolldata characteristics of Chongqing freeway, and then establishes the ACI algorithm,which is based on the toll data of freeway. Because the performance of ACI algorithm isaffected by the number of sample vehicles, the paper improves the ACI algorithm.Though the example, the improved ACI algorithm has a better performance.③In order to solve the problem of ineffective identification of abnormal trafficstate, which is based on the single data source, the paper establishes identification ofabnormal traffic state, which is based on data fusion. To solve the problem of algorithm fusion, the paper introduces the voting fusion method to identification of freeway trafficconditions, and establishes the ACI algorithm fusion method. The fusion method mainlyincludes three parts: the ACI algorithm based on vehicle sensor data, the ACI algorithmbased on toll data and fusion algorithm. Though the example, the method has a betterperformance, which is based on information fusion.