Research On Prediction Model Of Highway Traffic Flow

In order to effectively solve the problems of traffic congestion,intelligent transportation system came into being.Traffic flow prediction provides support and guarantee for its.How to obtain accurate and realtime traffic flow prediction results is an urgent problem for researchers.Therefore,the main research content of this paper is to establish a model to predict highway traffic flow.Through the comparative analysis of the prediction results of different models,the best prediction model is obtained,which can be used for traffic flow prediction.The main work is as follows:Firstly,the traffic flow data is analyzed,the characteristics of traffic flow data are mined,and the feature construction,extraction and selection are carried out through data preprocessing,which effectively makes up for the problem of single feature.Then,the prediction model based on extreme gradient boosting(XGBoost)is established,and the model is optimized by grid search CV method.The reliability of the model is verified by an example.Secondly,considering the factors of training time and prediction accuracy,the fruit fly optimization algorithm(FOA)is proposed to optimize the light gradient boosting machine(Light GBM),and a highspeed highway traffic flow prediction model based on FOA-Light GBM is established,and the traffic flow data on weekdays and holidays are predicted and analyzed.Thirdly,in order to further improve the accuracy of the prediction model,through the analysis of the temporal and spatial characteristics of traffic flow,this paper proposes to use the convolutional neural networks(CNN)to extract the spatiotemporal characteristics of traffic flow,and then combines with FOA to optimize the Light GBM model to establish a highway traffic flow prediction model based on CNN-FOA-Light GBM.The effectiveness of the model is verified by an example.Finally,the research results show that the real-time and accurate prediction of traffic flow can provide effective support and guarantee for its.