Exploring correlated parking–charging behaviours in electric vehicles: a data-driven study

Abstract

To enhance the management of parking–charging behaviours for electric vehicles (EVs) and promote the development of vehicle–grid interaction technology, the interrelation between parking and charging behaviours among EV users should be investigated further. This study, based in Changshu City, Suzhou, China, established a data linkage mechanism for parking–charging platforms and developed an EV parking–charging behaviour database, considering critical metrics like charging start time, initial and final state of charge, and charging duration. Employing the K–S test and K-means clustering methods, the diversity in parking–charging preferences between pure and plug-in hybrid EV users is explored. Results indicate that pure EVs’ parking–charging behaviours can be categorised into five distinct groups using a classification model, while those of plug-in hybrid EVs can be grouped into four categories. Both user groups include behaviours with low range anxiety, such as complete charging during special journeys, at the destination, or partial charging. Both groups also exhibit high-range-anxiety behaviours, with pure EV users favouring specific journey complete charging and plug-in hybrid EV users preferring complete charging. Notably, pure EV users also show a significant inclination towards nighttime complete charging. These insights are valuable for efficient planning and management of integrated EV facilities.