Research On Complex Road Network Model And Its Optimal Path Search Algorithm

With the acceleration of urbanization development, urban trafficproblems also appear more and more prominent,and intelligent transportationsystem is considered to be an important way to solve these problems. Roadnetwork model, the traffic flow prediction and path search algorithm as animportant part in intelligent transportation system, the their combination willgreatly improves the efficiency of the dynamic traffic navigation system.Thiswill effectively integrate resources to city traffic management system, solvethe problem of traffic congestion in big cities and improve the urban trafficmanagement level. This paper focus on road network model,traffic flowprediction method and path search algorithm of ITS.The main work can beincluded in the following several parts.First of all, a novel road-based model, based on the traffic networkfeatures and the human’s understanding to the road network, is constructed inthis paper. In the this model,the traffic and spatial attributes of the roads areexpressed in detail, and the traffic network is described by the roadintersection relationship.At the same time, a large section of the trafficinformation are stored in the road in order to reduce data storage redundancyand improve the network the efficiency of inquire.In addition,the model is alsoable to describe dynamic information through setting related variables.Second，the development process and development status of traffic flowin the futrue is smilar to development process and development status oftraffic flow in the past,so the traffic flow of some road sections of the next moment is related to traffic flow of past,and the traffic flow has certainsimilarities and trend.Therefore, analysising traffic flow data using the theoryand method of time series,a traffic flow prediction method based on singlevariable time series is builded up. At the same time, to improve the predictionaccuracy, traffic flow data are denoised by using kalman filtering.Third, on the basis of the RBM model,an improved A*aglorithm isproposed.At the same time,according to the fearure of traffic flow,the changeof road traffic situation in the period from origin to destination is integrated inpath search process of improved A*algorithm by the estimated transportationcost,thus a highly efficient dynamic improved A*algorithm for pathoptimization is obtained.In addtion, an enhanced ant colony algorithm basedon novel road model is proposed for improving the convergence speed.Finally, with the MapInfo building road network space model and withaccess database storing the road network data,a the traffic network dynamicpath optimization software is progromed by VB. Then the traffic network ofBeijing within the five rings is regarded as search examples.The examplesshow that the proposed model can meet the needs of practical application andthe path search algorithm based on traffic flow predictiion can improve theefficiency for drivers’ travelling.