Grid-ECO: Grid aware electric vehicle charging stations placement optimizer
Abstract
The rapid growth of electric vehicles requires expanding public charging stations, but adding stations to existing grids can overload transformers and cause voltage drops, leading to expensive upgrades. We formulate and solve NP-hard Mixed-Integer Nonlinear Program (MINLP) to optimize charging station placement and sizing by maximizing number of chargers ensuring grid feasibility, budget constraints, and demand fulfillment. We reformulate the MINLP problem into Mixed-Integer Bilinear Program (MIBLP) that the Gurobi solver can solve to near 0% optimality gap if good initial conditions and bounds are available. We applied McCormick relaxation on MIBLP and bound tightening to get good initial condition.
Primary Faculty Mentor Name
Amritanshu Pandey
Status
Graduate
Student College
College of Engineering and Mathematical Sciences
Program/Major
Electrical Engineering
Primary Research Category
Engineering and Math Science
Grid-ECO: Grid aware electric vehicle charging stations placement optimizer
The rapid growth of electric vehicles requires expanding public charging stations, but adding stations to existing grids can overload transformers and cause voltage drops, leading to expensive upgrades. We formulate and solve NP-hard Mixed-Integer Nonlinear Program (MINLP) to optimize charging station placement and sizing by maximizing number of chargers ensuring grid feasibility, budget constraints, and demand fulfillment. We reformulate the MINLP problem into Mixed-Integer Bilinear Program (MIBLP) that the Gurobi solver can solve to near 0% optimality gap if good initial conditions and bounds are available. We applied McCormick relaxation on MIBLP and bound tightening to get good initial condition.
