Resilience-Oriented Computationally Intensive Operation of Renewable-Powered Reconfigurable Microgrids: A Matheuristic Approach

Abstract

This paper addresses the challenge of achieving both sustainability and resilience in renewable-powered reconfigurable microgrids (RMGs) while managing computational burdens. The Green Independence Performance Index (GIPI) is introduced as a novel resilience metric designed to maximize microgrid independence from the utility grid while avoiding renewable energy curtailment. A multi-objective, nonlinear, and non-convex optimization model is developed to simultaneously minimize total operational costs and maximize GIPI. To enhance tractability and guarantee solution optimality, the problem is reformulated as a mixed-integer linear programming (MILP) model using piecewise linearization and approximation techniques. To further reduce computational burden, a custom matheuristic algorithm is introduced, which synergistically combines classical MILP optimization with a neighborhood-structure-based heuristic. Numerical experiments on the enhanced IEEE 33-bus benchmark system demonstrate that the proposed GIPI eliminates renewable energy curtailment, unlike the IPI which curtailed 13.58% of photovoltaic and 25.09% of wind generation, while the proposed matheuristic reduces computational time from 1,629.74 s (CPLEX) to only 103.32 s and maintains near-optimal operating costs.