Traffic Congestion Prediction Model Based On Time Series Association Mining Rules

As the level of urbanization continues to increase,the problem of traffic congestion becomes more and more serious.If we can predict the traffic flow through certain means,the urban traffic pressure can be alleviated.In the urban road network,traffic congestion presents the characteristic of radiating to the surroundings.In this paper,the congestion conduction is modeled from the two dimensions of time and space.Mainly studied the traffic congestion prediction method and congestion conduction law.A mining model of time series association rules based on genetic network programming algorithm(GNP)is proposed.By mining the congestion transmission rules between urban road networks,the traffic state of future roads can be predicted.This research first introduces the relevant theory,classification basis and classic algorithm of association rule mining algorithm.At the same time,it explains the time series association rules and its application in the field of transportation.Next,the article introduces related concepts of genetic network programming algorithms(GNP).The basic structure,evolutionary process and genetic operators involved are analyzed in detail.By comparing the GNP algorithm with the classic association rule mining algorithm,it provides a theoretical basis for data modeling.Then,the article establishes a GNP-based association rule mining model.In the study,the operation process of the algorithm,the principle of rule generation are explained in detail,and the association rules of traffic jam are defined.In addition,a method for obtaining traffic state data is introduced in this chapter.Collect the traffic situation data in the open platform of Amap by means of crawling.This method provides data support for empirical analysis.Next,an empirical analysis of road traffic situation data within the Third Ring Road of Beijing.From the wide-area and local perspectives,analyze the model prediction effect and congestion conduction law.By comparing with the neural network-based congestion prediction model,the model established in this research can better predict the future traffic congestion.Compared with the black box prediction process of neural networks,this study can better explain the prediction results of the model based on the excavated road congestion transmission information.The traffic congestion conduction law extracted by this research can not only predict the future traffic state.Establish a congestion warning and prevention mechanism to reduce the incidence of road congestion and improve the efficiency of traffic travel.Moreover,in this study,through an intuitive display of the congestion conduction law in local areas,taking Guomao Bridge,Yuquanying Bridge and Beijing West Railway Station as examples,this paper analyzes the impact of different road structures and road environments on the traffic congestion conduction speed and response time.By identifying the imperfect parts of the road plan,relevant suggestions for improvement planning are proposed.Help traffic managers and city builders to better plan and design traffic road networks,and promote the development of service technology in the field of intelligent transportation towards intelligence,diversification and specialization.