Study On The Speeding Regularity Of Commercial Vehicles Based On Clustering And OLAP

With the highly development of road traffic and the increase of commercial vehicles, traffic safety becomes more and more severe. It has been proved that speeding is one of the main reasons for traffic accidents. However, long low-level informationization in operation department, the lack of data and advanced technology backup has resulted in striking deficiency in safety management.Now there are large quantities of data from commercial vehicles equipped with GPS receiver with the application of GPS in transportation management, which supplys the basis for the administration department's obtaining of speeding regularity. And meantime the development of information technology provides for a new approach to analyse the speeding based on data warehouse, OLAP and clustering. Therefore, it is necessary to obtain the regularity of commercial vehicles speeding through the GPS data based on clustering and OLAP to help safety management and improve the level of management.The research described in this paper focuses on further analysis of overspeed data from GPS equipped commercial vehicles. And then a general solution to detect the regularity of speeding is presented based on the application of clustering and OLAP technology.Initially, current situation of the vehicle speeding and some related theories are discussed in the paper. Then a general solution is presented. Considering the difficulties to acquire overspeed black spots of commercial vehicles, a density-based clustering algorithm is proposed to search for the spots where the speeding usually happens in high frequency based on the deficiency of the original one. To simplify the data structure building process and reduce the memory space it occupied, the traditional algorithm is improved with the adjacency list replaced the R*-Tree. On the basis, this paper presents that it is efficient to apply OLAP technology to the multi-analysis of commercial vehicles speeding according to the requirements of transportation management sector. Several key technologies are resolved during the process of its realizition, including finishing data pre-processing, establishing data cube based on Analysis Services, accessing data cube through MDX and ADO MD and doing research on OLAP multi-table-join algorithms to improve the performance of query.Finally, the proposed solution is tested by the vehicle alarm data from GPS monitoring system in Chongqing. The results indicate that the improved clustering algorithm outperforms the original one in terms of efficiency and the application of OLAP to the analysis of speeding can efficiently implement multi-dimensional analysis and visualization.