Automated ARIMA Model Construction for Dynamic Vehicle GPS Location Prediction

Abstract

Many applications in intelligent transportation systems are demanding an accurate vehicle GPS location prediction. In this study, we satisfy this demand by designing an automated GPS location prediction system based on the well known traditional Auto-Regressive Integrated Moving Average (ARIMA). To increase the proposed model accuracy, make it dynamic, and reduce its execution time, the traditional ARIMA model has been modified extensively. To perform GPS location prediction, the proposed model depends on a given vehicle previous locations to predict the vehicle future location. To make it dynamic, the proposed model is designed to regenerate all its parameters every period and only consider a specified window in the history. The proposed model is evaluated based on real vehicle dataset traces that we recorded using an app on a smart phone. The results show that the proposed framework can generate ARIMA models that can predict the GPS location of a vehicle in the future accurately and with a reasonable execution time. If the application needs harder deadline (shorter deadline), we propose to use the last ready model to predict the next vehicle’s GPS location.