Detection Of Traffic Congestion Regions Based On Bus Trajectory Data

With rapid development population increased of the city,traffic congestion is the common problem of developed big city,which not only affects the quality of people travelling,but also become one of the main obstacles for future development of the city.The essential cause of traffic congestion is the existing structure of the transportation system can not meet the growing travel needs.So it needs reasonable traffic planning to improve the existing traffic structure.First of all to do is recognizing the existence of regional traffic congestion.This paper uses data mining technology to explore and analysis traffic congestion areas using real data,which provides data base for traffic planning policy setting.First of all,this paper describes the urban traffic conditions by using urban bus trajectory data and bus line-site data,which including dividing data by time and space,feature extraction and data matrix construction.After that,this paper uses a method of anomaly detection which is based on global outlier factor to identify the bottleneck of traffic congestion.Finally,in order to analysis combined with travel behavior,regional properties and city development,this paper proposes a transit oriented city region division method.It uses public transport sites for regional traffic center clustering and Voronoi plane partition method to identify the traffic zone boundary.And combine road traffic congestion detection results with the regions to explore traffic condition of each sub regions.In this paper,we use the data of bus to detect traffic congestion areas in Hangzhou City.We present the results with thermodynamic diagram and analyze using real traffic planning,social events and regional social attributes.We use the ten most congested roads released by integrated transport research center in Hangzhou to do quantitative verification of our detection results.The two ways both verify the effectiveness and superiority of our proposed detection method of congestion in the city area.