Enhanced Particle Swarm Optimisation for Optimal Economic Dispatch of a Standalone Renewable Microgrid

Abstract

Text size

Optimal energy management in renewable-based standalone microgrids is a complex nonlinear optimization problem, particularly under Malaysian operating conditions where solar photovoltaic (PV) and wind turbine (WT) generation are affected by weather variability. This study proposes an Enhanced Particle Swarm Optimization (E-PSO) framework for the optimal economic dispatch of an islanded microgrid consisting of a 200 kWp PV system, a 100 kW wind turbine, a 150 kW diesel generator (DG), and a 300 kWh battery energy storage system (BESS). The objective function minimizes the total daily operating cost by accounting for diesel fuel costs, CO₂ emission penalties, battery degradation costs, and energy not served (CENS) costs. Minimum up-and-down time constraints for diesel generators are handled using binary commitment variables and a post-update repair mechanism. At the same time, the power balance formulation includes energy not served and renewable curtailment to ensure operational feasibility. Simulation results over a 24-hour scheduling horizon show that the proposed E-PSO achieves the lowest operating cost of USD 319.50/day, reducing costs by 27.4% compared with the load-following strategy (LFS) and 15.1% compared with the genetic algorithm (GA). Renewable penetration reaches 59.6%, higher than 46.8% under LFS. Statistical evaluation over 30 independent runs gives a standard deviation of USD 1.72, indicating stable convergence. Sensitivity analysis under varying solar irradiance, wind speed, BESS capacity, and carbon penalty confirms the robustness of the proposed framework for day-ahead operation of standalone renewable microgrids in Malaysia.