Analyzing overspeed regularities of commercial vehicles based on data mining

This paper focused on analyzing the over-speed regularities of commercial vehicles which canpsilat be well acquired nowadays. For the real-time monitoring data from the GPS receivers equipped on about 5000 commercial vehicles in Chongqing City, China were available, the data mining technology was introduced to analyze the over-speed regularities based on the huge data set. Firstly a map-matching algorithm was proposed to preprocess the spatial attribute data. And then a four-dimensional data warehouse model was established. On the basis, pattern mining of data was realized by FP tree-based algorithm of association rules and tendency analysis method. The results show that the obtained patterns are effective enough to provide valid decision-making support for transportation safety management.