Research On Traffic Condition Identification And Prediction Of Urban Ring Road

With the rapid development of national economy in China, the number of motor vehicles increases year by year, which makes the traffic situation of urban ring roads worse and worse. As the "main skeleton" of urban road traffic network system, the traffic condition of urban ring road will affect the whole urban road traffic network system. Therefore, solving traffic problems of urban ring road is equal to improving the traffic condition of the whole urban road traffic network system and it will be of great significance.As a component part of ITS, if traffic condition identification and prediction system can provide real-time and accurate traffic condition information to road users, it will greatly improve the travel efficiency and save travel time. At the same time, it will help traffic manager make decision more convenient and make traffic control and management measures more rational and effective. It will be of great significance to alleviate traffic congestion.Therefore, this article focuses on traffic condition identification and prediction of urban ring road. First, the paper describes research background and meaning of traffic condition identification and prediction of urban ring road, and reviews research on traffic condition identification and prediction. Then, the characteristics of urban ring road are analyzed to explain the difference between urban ring road, highway and general urban road. Traffic flow parameters and features of urban ring road are introduced and the relationship between traffic flow parameters is analyzed. At the same time, the paper analyzes the influence factors of traffic condition of urban ring road.Base on the analysis of relationship between traffic flow parameters and influence factors of traffic condition, the paper put forwards traffic condition identification model based on different space and different influence factors. The traffic identification model of different space is based on fuzzy c-means clustering (FCM) and grey clustering. First, boundary of different traffic conditions is divided based on FCM method and threshold value of different traffic conditions is calculated. Then traffic conditions are identified based on grey clustering model. The model can avoid the clustering method trapping into a local optimal solution and decrease the subjectivity when setting grey clustering’s parameters. The traffic identification model of different influence factors is based on self-organizing maps network. The method doesn’t need external target output to look for the type in complex data.After finishing the traffic condition identification model, the corresponding traffic condition prediction system is designed. By the traffic condition prediction system, users can search traffic condition information on the internet. Finally, taking Fuzhou Second Rind Road for example, the traffic condition identification model is validated by computational analysis of traffic flow data, and the identification result is released through traffic condition prediction system.