Research On Probability Density Estimation And Prediction Method Of Road Traffic Density

With the continuous development of the social economy,the car ownership continues to rise,and traffic congestion has become a major problem in road traffic management.Traffic density is an important basic characteristic parameter in traffic flow theory,which can make a good evaluation of the service level of roads.The study of traffic density can provide theoretical basis for traffic regulation and solving road congestion problems.However,the current research on traffic flow density mostly focuses on the detection or estimation of traffic flow density data.Based on the knowledge of statistics and probability theory,the research on the probability density of traffic flow density and the analysis of its distribution characteristics are still few.This paper studies the probability density estimation and prediction of traffic density.The main work can be described as follows:For the original traffic flow density data obtained,there may be fault data such as missing and abnormal data.The threshold screening method is combined with the traffic flow theory screening method to identify the fault data and preprocess it with historical trend value replacement method.Through this preliminary work,the quality of the data is guaranteed and the accuracy of the research results is improved.A probability density estimation method based on Gaussian mixture model(GMM)for traffic flow density is designed.Firstly,the fuzzy C-means(FCM)algorithm is used to complete the parameter initialization of GMM,so as to speed up the convergence speed when the expected maximum(EM)algorithm solves the model parameters.Secondly,the EM algorithm is used to solve the GMM parameters.Then,we can get the probability density estimation result of traffic flow density according to the GMM.In this paper,based on this method,the probability density of traffic flow density on weekdays and weekends is estimated for four locations with different traffic conditions.The experimental results show that the GMM has higher accuracy than the conventional single probability distribution models.Based on the study of probability density estimation of traffic flow density,this paper further designs a method for predicting the probability density of traffic flow density.Firstly,the prediction model is constructed based on the idea of time series.Then the objective function of the prediction model is constructed based on KL divergence.Finally,the optimal parameter combination in the prediction model is obtained by Bayesian optimization algorithm.The experimental results show that the proposed prediction model can predict the probability density distribution of traffic flow density,whether it is for ordinary road sections or complex intersections.Through theoretical analysis and experimental verification,the probability density estimation and prediction method of traffic flow density designed in this paper is feasible and accurate,and can provide effective help for assessing road service level and traffic management regulation.