Multi-Objective Optimal Scheduling of a Grid-Connected Microgrid Using an Improved PSO

Abstract

Microgrid optimal scheduling is an effective way to reduce operation cost and environmental impact under high penetration of distributed generation. This paper proposes a multi-objective dispatch model for a grid-connected microgrid that simultaneously considers operating expenditure and emission treatment cost. The microgrid consists of photovoltaic (PV) units, wind turbines (WTs), a diesel generator (DE), a micro gas turbine (MT) and a battery energy storage system (BESS). Detailed models for these units are established, and the total cost, including fuel, operation-maintenance and pollutant treatment, is formulated as the objective function. To enhance global search capability and avoid premature convergence, an improved particle swarm optimization (IPSO) algorithm with dynamically adjusted inertia weight and learning factors is designed to solve the scheduling problem. Simulation results on a typical 24-h microgrid case demonstrate that the proposed model can effectively reduce both electricity cost and environmental cost, and the IPSO achieves lower objective value and better robustness compared with the conventional PSO algorithm.