Research On Urban Expressway Traffic State Identification Method Based On Cloud Model

With the rapid increase of traffic demand, traffic congestion is inevitable, restricting the socio-economic development seriously. In order to solve the traffic congestion problem and meet the rapid travel demands of the various groups within the city, the expressway network have been built in large cities. As the basic framework of urban transport system, expressway bears a large proportion of motor vehicle traffic. Because of the good accessibility and intensive ramp design of expressway, the expressway traffic status shows a trend of deteriorating year by year, the trend is mainly manifest as: increasing congestion intensity,creeping the scope of congestion and growing the time of congestion. As a result.expressway traffic efficiency have significantly reduced, missed its functional position seriously. Therefore, to improve the traffic operation efficiency of urban expressway has become the primary task of urban traffic management.Urban expressway traffic state has the strong random and non-linear characteristics, the identification process is very complex, to accurately extract real-time, reliable traffic status information is the premise of intelligent traffic management. The paper mainly focuses on the multi-source information fusion based on cloud model, and extracts the traffic information of expressway. The main researches in the paper are as follows:Firstly, the paper reviewed the current theories and methods of traffic status identification, the cloud model is used to identify the expressway traffic status. The paper filters the traffic state evaluation indexes, in order to avoid the interference of the second-order characteristics of the traffic on the recognition results, selects average speed and time occupancy as the evaluation indexes.Secondly, the paper analyzed data pre-processing methods to prepare to identify the model. Proposed method based on time series weights to complete the lost data. As the analysis of numerical example shows: The data filling method based on time series weight reflects the time-varying characteristics of traffic flow parameters, and the results are closer to the real value.Thirdly, the paper established an expressway traffic state identification method based on cloud model. Clustering historical data by K-means clustering method, using the backward cloud generator to obtain the initial cloud of traffic state, using trapezoid cloud to improve the accuracy of identification results; Ascertaining the dynamic weight of the evaluation index by the method of information entropy; finally output the traffic state and traffic state index.Finally, the paper used a specific case to verify the effect of the model. Based on the data of Chengdu Second Ring Road's microwave data to evaluate the traffic status,compared the identify results with speed threshold method and V / C ratio threshold method,analyzed these results to verify the validity of traffic state identification method based on cloud model. It is proved that traffic state identification method based on cloud model can improve the problem of single index threshold division method, and solved the problem that the traditional normal cloud model have not high accuracy in extreme cases, and it can reflect the real traffic status, have greater practicality and portability.