Research on Pre-processing Method of Travel Data Based on Mobile GPS: Three-step Data Pre-processing Method

Abstract

The collected travel data are accompanied by noise data and missing data. They seriously affect the analysis and use in transportation. Data processing is the foundation of data analysis and the most difficult step in data utilization. In order to realize the noise reduction and interpolation of the data, we creatively adopt the three-step method of data segmentation and deductive reasoning, the verification of Spearman's correlation coefficient and the discrimination of the distance of triangle bottom to height, which effectively repairs the abnormal data in this article. Through the processing of 78 travel data, the transfer point extraction, data interpolation and drift correction are realized. The results have high reliability. This method can be programmed and used for batch data processing, which will save considerable investment in the field of transportation survey.