Representative Charging Patterns for Electric Vehicles Using Image Clustering Algorithms

Ahstract-With the penetration of electric vehicles, concerns about operational efficiency are gradually increasing. However, due to the absence of public data on electric vehicles, indirect research has been conducted through the configuration of a simu-lation environment. Additionally, there is a lack of research using unstructured field data that was gathered from electric vehicles. To solve this problem, this paper proposes a method of deriving representative charging patterns through a newly proposed data pre-processing technique using field electric vehicle data collected by LG B-life care. By applying the proposed method, we confirm the possibility of deriving six representative charging patterns. Furthermore, by analyzing the fuel consumption of vehicles for each pattern, we show that it is possible to increase operating efficiency by more than 3% by influencing the driver's charging pattern.