A parallel clustering and test partitioning techniques based mining trajectory algorithm for moving objects

In recent years, the intelligent transportation system has been widely used to deal with traffic problems. The analysis of traffic incident is important in intelligent transportation field, and gathering patterns can model various traffic incidents. However, with the increasing amount of moving trajectory data, the traditional mining algorithms of gathering patterns cannot effectively analyze trajectory data. In this paper, we propose a parallel algorithm RDD-Gathering to discover the gathering patterns in massive trajectory data. Based on the algorithm, we further design a system framework of traffic incident analysis and prediction, which can realize the prediction of the abnormal traffic events. Finally, the accuracy and efficiency of the proposed algorithms are validated by extensive experiments based on a real trajectory dataset, and the results of experiments show that the proposed method can effectively improve the efficiency of gathering retrieval.