Data-Driven Clustering Analysis for Representative Electric Vehicle Charging Profile in South Korea

Area of Science:

• Electric vehicle technology
• Energy systems analysis
• Data science

Background:

• Increasing electric vehicle (EV) adoption necessitates understanding their operation for grid stability and battery health.
• Variability in user charging habits complicates unified analysis of EV operational data.

Purpose of the Study:

• To propose a framework for analyzing EV charging characteristics using actual field data.
• To model individual EV operation profiles in a unified format.
• To identify representative EV charging patterns.

Main Methods:

• Modeling individual EV operation profiles using probability distributions of charging start and end states of charge (SoCs).
• Employing hierarchical clustering analysis to derive representative charging profiles.
• Utilizing large-scale, vehicle-specific EV data from South Korea.

Main Results:

• A framework was developed to capture EV charging characteristics based on charge-discharge areas.
• EV operation profiles were modeled using SoC probability distributions.
• Hierarchical clustering identified seven distinct representative EV charging profiles.

Conclusions:

• The proposed framework effectively unifies diverse EV charging data.
• Seven representative charging profiles were identified, simplifying the analysis of EV battery charge-discharge behavior.
• This classification aids in understanding and managing EV integration into the power grid.