Freeway Traffic State Classification And Visualization Analysis Based On Toll Station Data

Based on big data mining and Intelligent Traffic System(ITS)technology,the effective identification and visualization of traffic operation status has become an important methods to understand the abnormal operation of traffic.Compared with other road transportation modes,freeways have the characteristics of fast flow rate and large time and space span.The research on real-time discriminant classification of freeway traffic status is great significance to improve road traffic efficiency and reduce property loss.At present,the relevant research data sources are diverse,the data structure is complex,and the computational analysis is difficult.Using a vested data as the data foundation can effectively reduce sensor placement and calculation costs.At the same time,how to select a set of high precision and quantifiable index system,and based on this system,given the classification standard of road state classification,to find the quantitative boundary of classification has been the focus of research in this field for a long time.Based on the above questions,this paper constructs the data model of freeway scene and the traffic state discriminant multi-index parameter system by using the toll station data that is set on the boundary of the urban area to record the passing vehicles in detail.According to the characteristics of freeway operation,the classification criteria of time-level classification and space-time feature are constructed respectively,and the fuzzy set theory is introduced to solve the problem of fuzzy classification boundary,and realize the classification of freeway traffic level.Finally,the visual analysis of characteristics of freeway traffic flow and operating state is realized by using the method of statistical chart visualization and space-time feature visualization.The specific research work is as follows:(1)Analyze the data characteristics of toll stations and the characteristics of freeway traffic,mine real-time dynamic traffic information from static two-dimensional traffic data,build a data model based on the space-time association of people,roads and toll stations in freeway scenes,and use the Column Union Row(CUR)decomposition and the wavelet transform method for feature analysis and preprocessing of freeway traffic data,which laid a data foundation for subsequent research.(2)The factors affecting traffic flow state are excavated from the perspective of traditional time series model and external influence factor model,and the discriminant system of multi-index parameters is constructed.The gradient tree decision tree prediction algorithm,which can process discrete values and continuous values at the same time,proves the availability of the index system,makes up for the problems of predecessors’ influencing the influence mechanism of traffic flow,and improves the accuracy and completeness of traffic state classification.(3)Based on the classification theory of traffic state classification,the classification criteria are established from the perspective of road traffic level and space-time characteristics,and the space-time classification system is described in detail from the perspective of data features and classic cases.Finally,the fuzzy set theory is introduced for the boundary fuzzy problem faced by the traffic state classification.The fuzzy C-Means Algorithm(FCM)clustering method is used to classify the training sample data to obtain the cluster centers of each category.The automatic discrimination applicable to the traffic state of the freeway is realized.(4)In order to realize the feature characteristics of freeway data and the extraction of feature association structure,the traditional multivariate data visualization method and GIS space-time visualization technology are used to realize the visual analysis of freeway traffic operation mode.Based on the state classification example of the characteristic area freeway network,the paper validates the paper algorithm and discusses the influence of weather elements on the flow operation mode.