Research On The Method Of Traffic State Recognition Of Urban Expressway Based On Machine Learning

The real-time discrimination of urban expressway traffic state is an important reference for traffic management departments to make decisions.At present,the research of traffic state discrimination of urban expressway always put congestion state into traffic abnormal events,or take a certain threshold of traffic parameters as the criterion of traffic state discrimination.The accuracy of this method is not high and it can not be used for real-time discrimination.In this paper,a traffic state discrimination model of urban expressway is built by using fuzzy clustering method and support vector organization based on one-day historical data of a road section,which improves the accuracy and speed of discrimination.Firstly,the paper introduces the traffic flow parameters,expounds the relationship among traffic flow,speed and density,determines the characteristic parameters(flow,speed and occupancy)of traffic flow used in this study,and determines the division standard of traffic state of parametric expressway.Secondly,genetic algorithm and fuzzy c-means clustering algorithm are introduced,and the process of traffic state discrimination using fuzzy c-means clustering algorithm is described.Then,in view of the shortcomings of fuzzy c-means clustering,this paper improves the fuzzy c-means clustering algorithm,inserts the fuzzy c-means clustering into the process of genetic algorithm,and proposes a parallel genetic fuzzy c-means clustering algorithm,which effectively solves the problem that the fuzzy c-means clustering algorithm is sensitive to the initial value.Because there is a large amount of calculation in the method of traffic state discrimination based on Euclidean distance,the classification model of traffic state is established by using support vector machine,and the parameters of support vector machine are optimized in three ways:grid search method,genetic algorithm and particle swarm optimization algorithm,so as to obtain the parameter group that can make the training model reach the maximum accuracy.Finally,the model is verified by the measured data.The parallel genetic fuzzy c-means clustering and the original fuzzy c-means clustering are compared in convergence speed,sensitivity to initial value and clustering effectiveness.The results show that the selection of initial value has no effect on parallel genetic fuzzy clustering,and each iteration can converge to the global minimum.The number of iterations is very small.The clustering efficiency is also higher than the original fuzzy c-means clustering.This method provides a good foundation for the subsequent training of SVM and saves a lot of time.