Research On Prediction And Visualization Of Urban Expressway Road Conditions Based On Travel Planning

With the development of urbanization and the improvement of residents’ living standard,the year-on-year growth rate of motor vehicle ownership in urban areas is much higher than that of urban road length,and the contradiction between road supply and travel demand is gradually increasing.In order to improve the travel efficiency and travel experience in the limited road carrying capacity,the research on the future road condition prediction is essential.Relying on the National Nature Fund,Professor Xu Weixiang’s team is trying to make the road situation transparent in the future.Under the guidance of the WWF project,Professor Xu Weixiang’s team has done a series of research work on the traffic status of the road network and the technology of car networking,and put forward the travel plan.Based on about 15 GB of data in 4 modules,such as travel planning data,traffic speed data,road GPS data and road basic information data,this paper first preprocesses the data,analyzes the correlation of speed data,then fits the model based on the travel plan traffic data and the real road speed data,then predicts the future road speed,and finally visualizes the road traffic status.The main research includes the following four parts:(1)Based on the idea of travel planning,this article proposes a method of judging the range of latitude and longitude to convert Baidu map user query data into road section travel plan data,and then into road flow information.At the same time,the correlation analysis of the road real speed data is carried out from the spatial dimension and different time latitudes,and the Person correlation coefficient is used to prove the spatio-temporal correlation of the speed data.(2)Based on the three parameters of speed,flow and density of traffic flow and the basic graph model,a speed-flow model based on the travel plan flow data and the road real speed data is proposed.By using Origin,the data is treated with low-pass filtering,and the fitting accuracy of different models before and after filtering is compared(3)By comparing and analyzing the basic ARIMA model,the BP neural network based on the artificial neural network model and the LSTM model based on deep learning,the LSTM was selected to establish a model for predicting the average travel speed of the road in the future.The model was trained and verified using actual data,and the prediction results were analyzed.(4)According to the traffic speed data,the road traffic status is visually displayed in the form of heat map through the Map Lab platform.By studying the travel planning data,this paper makes some attempts in the transparency of road traffic conditions,hoping to help alleviate the pressure of urban traffic and the development of intelligent traffic in the future.