Search

Users of electric vehicles (EVs) have different preferences for the temporal charging patterns. The analysis of temporal charging patterns can aid in the decision‐making of EV‐related stakeholders, including policymakers, grid operators, and automobile companies. Herein, a recognition and factor analysis framework of temporal charging patterns for EVs’ usage improvement is proposed. A Gaussian mixture model (GMM) is used to obtain temporal charging patterns by considering characteristics related to the charging session and time of parking period. A multinomial logit model is used for the factor analysis of the temporal charging patterns. An experiment is conducted with the proposed framework using data of private electric vehicles (PEVs) and carsharing electric vehicles (CSEVs) collected between 2018 and 2019 in China. The data are statistically analyzed, and then six types of temporal charging patterns are obtained by the GMM. A comparison of the pattern distributions over time shows that they are obviously different between PEVs and CSEVs. The analysis of the patterns’ influencing factors of PEVs and CSEVs shows that the PEVs have more influencing factors than the CSEVs. Herein, a recognition and factor analysis framework of temporal charging patterns is proposed. The Gaussian mixture model is used to obtain temporal charging patterns. The multinomial logit model is used for factor analysis. The proposed framework is verified experimentally using data of private electric vehicles and carsharing electric vehicles in China collected between 2018 and 2019.