LAUSR

As one of important implications on electric vehicles, driving cycles are recognized as essential components for evaluating the comprehensive performances and they have drawn much attention for research. Currently, driving cycles are constructed specifiedly in international standards based on local traffic conditions. However, without consideration of the private driving habits, unproper cycles lead to the imprecision on predicting the remaining useful life or estimating states. Herein, a novel methodology based on Markov chain and Monte Carlo method is developed to extract the personal driving characteristics as the elements of divided kinematic fragments. Principal component analysis is adopted to address the high-dimensional parameter vector, and cluster is used to classify the kinematic fragments. The statistics analysis demonstrates that the processed database exhibits great consistency with our developed driving cycle compared against original database, where temperature, state-of-charge and consistency are utilized to describe the personal patterns. Moreover, by using the operational driving data, the developed driving cycle is comparable against other driving cycles, which exhibits good performance. Overall, the presented driving cycle of electric vehicle can be considered as an effective way in evaluating the private driving habits, predicting the battery states and other related applications. The method may be promoted for future better energy management on electric vehicles owing to the promotion of connected and autonomous vehicles.