Optimal Scheduling of Microgrid Using Constrained Multi-Objective Optimization with Additional Objectives

To address the issues in the constrained multi-objective optimal scheduling problem of microgrids, such as encountering infeasible solutions and a low proportion of feasible solutions, we propose a constrained multi-objective optimization algorithm assisted by an additional objective function. Three swarms in a push-and-pull co-evolutionary framework are constructed: the first swarm optimizes the original constrained problem, the second swarm optimizes constraint violation as an additional objective to effectively increase the proportion of feasible solutions, and the third swarm uses push-pull search and improved constraint relaxation techniques to transform high-quality infeasible solutions into feasible ones during the search process. This helps the populations overcome infeasibility barriers and escape local optima. All feasible solutions found by the three populations during the search process are stored in a separate archive to ensure the feasibility and diversity distribution of the solutions. The proposed constrained multi-objective optimization algorithm is applied to optimize the scheduling of microgrid equipment by establishing a microgrid optimization model for combined cooling, heating, and power supply, with system operating cost and environmental management cost as the optimization objectives, and considering the constraints in actual operation. The results demonstrate that the proposed algorithm can reasonably adjust the output of each distributed power source and improve the flexibility and economy of microgrid optimization.