Probabilistic operational management of a renewable-based microgrid considering uncertainties using the self-adaptive gravitational search algorithm

Increasing uncertainties in electricity prices, load demand, and renewable energy generation pose significant challenges for optimal microgrid operation in deregulated electricity markets. This paper proposes a self-adaptive Gravitational Search Algorithm (SGSA), which enhances the standard GSA by incorporating a self-adaptive mutation operator with two movement strategies to mitigate premature convergence and improve solution quality. To model uncertainties in load demand, market prices, and renewable outputs, the 2 m-Point Estimation Method (PEM) is employed as a computationally efficient alternative to conventional stochastic approaches. The proposed SGSA-PEM framework is applied to a low-voltage microgrid consisting of microturbines, phosphoric acid fuel cells, photovoltaic units, wind turbines, and battery storage. Simulation results indicate that the integration of battery storage reduces the total generation cost by up to 49.7%, while renewable energy penetration increases by approximately 10% during peak demand periods. Furthermore, comparative analysis shows that SGSA achieves lower operating costs and converges about 25% faster than standard GSA and Particle Swarm Optimization (PSO). The results confirm that the proposed framework provides a computationally efficient and robust solution for probabilistic microgrid energy management under uncertainty.