In the electric vehicle remote monitoring and intelligent control project, SIM800C module every two seconds to upload a longitude and latitude data to the server, the server location by receiving data, realtime monitoring and intelligent control purposes. However, when the GPS terminal into the location of blind spots, or the occurrence of the stars and the number of search satellites less than four cases, will lead to the location of the data is not accurate enough, resulting in data drift phenomenon. In this paper, the localization factor algorithm is improved and compared with the Euclidean geometry method and parabolic interpolation method. The experiment results show that the improved method can improve the accuracy of GPS data to a certain extent.