Algorithm of Vehicle’s Data Cleaning and Monitoring

The aim of this paper is to screen out abnormal data which is caused by human factors and natural factors. We proposed an algorithm of vehicle's data cleaning and monitoring. First, the valid data is filtered out by an improved DBSCAN method, through data analysis, and then get the threshold range. After that, it screens out the abnormal data in the valid data through the threshold range. Finally, the abnormal data is classified and counted according to the factors which was stipulated by the enterprise. The results show that the proposed algorithm can simpler and faster to process the abnormal data than the other similar algorithm.