Research On Traffic State Discrimination Method Based On Fuzzy Clustering

In recent years,with the development of the "Powerful Country in Transportation" strategy,more and more traveling modes have been selected,and requirements for traveling have become increasingly high.In the context of intelligent transportation,accurately obtaining the operating status of road traffic networks without adding road infrastructure is one of the main research contents of intelligent transportation systems(ITS).Establishing an accurate and real-time traffic state discrimination model not only helps managers grasp the overall traffic situation in a timely manner,but also facilitates traffic management departments to monitor road network traffic conditions in real time.This paper conducts research on the multi-time-scale characteristics analysis and parameter models of traffic flow,WKFCM algorithm and its optimization for traffic state discrimination.The main work is as follows:1.The multi-time-scale characteristics analysis and parameter model of traffic flow is studied.Aiming at the problem that the classic single-segment function model cannot accurately characterize the relationship of traffic flow parameters under different traffic conditions,a two-stage traffic flow parameter relationship model is proposed.First,the Lempel-Ziv algorithm in symbolic dynamics is used to calculate the time series complexity of traffic flow at different time scales.Second,Based on the correlation theory analysis method,the correlation of traffic flow in time and space is explored.Last,the traffic flow data collected by the PeMS system is used for characteristic analysis and parameter relationship model fitting.2.Research on traffic state discrimination method based on WKFCM.For the KFCM algorithm used for traffic state discrimination,there are problems of unreasonable metric function,a traffic state discrimination method based on WKFCM is proposed.At the same time,a traffic state discrimination method based on WKFCM is given,and the traffic state data is collected for clustering analysis of the traffic state.Finally,the false positive rate cross estimation method is used to evaluate the accuracy of state discrimination.3.Optimized WKFCM traffic state discrimination model based on improved adaptive GA.Aiming at the problem that the fuzzy C-means algorithm is sensitive to the initial cluster center and is easy to fall into a local optimum,an adaptive GA-optimized WKFCM traffic state discrimination model is proposed.Firstly,an adaptive genetic operator is designed,the initial clustering center is optimized with adaptive GA and the WKFCM algorithm is combined,and then the traffic flow data is used for cluster analysis and status discrimination.Finally,a comparative analysis with existing algorithms shows that the model in this paper has higher accuracy for traffic state discrimination.