Incorporating big data analysis in speed profile classification for range estimation

Incorporation of data from multiple resources and various structures is necessary for accurate estimation of the driving range for electric vehicles. In addition to the parameters of the vehicle model, states of the battery, weather information, and road grade, the driving behavior of the driver in different regions is a critical factor in predicting the speed/acceleration profile of the vehicle. Following our previously proposed big data analysis framework for range estimation, in this paper we implement and compare different techniques for speed profile generation. Moreover we add the big data analysis classification results to especially improve the performance of the Markov Chain approach. The quantitative results show the significant influence of considering the big data analysis results on range estimation.