EV Charging Station Siting and Mechanism Design
Problem and Approach
Public preferences expressed in natural language are difficult to incorporate directly into infrastructure planning. The Multilevel Joint Auction (MJA) framework combines LLM semantic parsing, mixed-integer linear programming (MILP), and mechanism design to map preferences into spatial demand and candidate sites, jointly optimizing charging-station siting and resource allocation.
My Contribution
I am the second author of the related paper. I contributed to mechanism modeling, optimization, experimental design, result analysis, and writing, including the conversion of natural-language preferences into structured planning inputs.
Current Status
The paper is at second-round revision at Applied Energy.