Research On Traffic Operation Situation Prediction Method Based On Online Map Data Mining

The identification and prediction of urban traffic operation status is the theoretical premise of urban traffic congestion management.In the existing research,the basic data is not easy to obtain,and the timeliness of the data is not taken into account well.It is difficult to identify the real-time traffic operation status.The evaluation of the overall congestion degree of the road network lacks a more intuitive output,and at the same time,it seldom provides quantitative information for individual travelers.On these grounds,this paper explores the law of urban traffic congestion based on real-time online map data,and constructs a real-time identification and short-term prediction method for urban traffic operation status.Firstly,the online map data types used in the papers are introduced,namely the online map heat map and the road travel delay index,and the methods and processing methods of the two types of data and the corresponding programming flow are given.Analyze the impact of urban land use and road network structure on urban traffic operation status,and establish a quantitative model of regional comprehensive thermal intensity.Using the acquired online map data,the trend of the interval time heat map data and the road travel delay index,and the relationship between the two types of data at continuous time points are analyzed.From the perspective of data analysis,the availability of two types of online map data in the research field of urban traffic operation state identification and prediction is proved.Secondly,based on the road travel delay index,the characteristics of road network system congestion are analyzed from the perspectives of disseminated road network system congestion and non-propagating road network system congestion.Combined with this characteristics,the collection of road network system congestion sections is defined,and the quantitative model of the road network overall system congestion and the fixed travel path system congestion is established.At the same time,the maximum road network congestion degree and path congestion degree are taken as the comparison objects of real-time identification,and the real-time traffic operation state identification method for traffic managers and travelers is constructed,and the specific algorithm is given.Taking Yuzhong District as an example,the MATLAB program is used to solve the problem.It is proved that the identification method can vertically compare the realtime traffic operation state of the road network or the travel path with its corresponding limit state,realizing the real-time identification of the traffic operation state.For the traffic manager or travelers provide important quantitative information on traffic operations.Finally,redefining the link units on different grades of roads.Based on the plume model,the method of determining the limit distance of the congestion spread of the link units is proposed,and the traffic influence range of the link unit is analyzed.The BP neural network model is used to predict the traffic operation situation of the fixed travel route and the road network: For the prediction of a fixed travel path,mainly based on the trend of the sections travel delay index and the comprehensive thermodynamic degree within the traffic impact range,and considering the time span of the traveler to each section,the two kinds of prediction ideas of the path start section travel delay index prediction and non-initial section travel delay index cross-time prediction are designed.For the prediction of the road network,according to the trend of the overall system congestion degree and comprehensive thermodynamic degree,directly predict the system congestion level of the road network as a whole in the future.Taking Yuzhong District as an example,the BP neural network toolbox of MATLAB software is used to prove that the prediction method has high precision.