A HYBRID METAHEURISTIC-SMC FRAMEWORK FOR SMART NANO- GRID CONTROL AND OPTIMIZATION

Abstract

A nano-grid represents a compact, self-sufficient power system that integrates both renewable and conventional energy sources to ensure a continuous electricity supply. In this study, the nano-grid incorporates a photovoltaic (PV) array, a wind turbine, and a fuel cell as its primary power generation units. To maintain stability in both active and reactive power outputs, various control strategies are employed, including integral (I), proportional-integral (PI), proportional-derivative (PD), proportional-integral-derivative (PID), fractional-order PI (FOPI), fractional- order PID (FOPID), and sliding mode control (SMC). To enhance the performance of these control methods, advanced optimization algorithms, namely Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), have been applied. These algorithms are guided by the integral square error (ISE) criterion, which serves as the objective function during optimization. Comparative assessments, using both graphical plots and tabulated data, were conducted to evaluate each controller’s efficiency and to determine the optimal configuration. Among all the tested controllers, SMC demonstrated the most effective performance in terms of maintaining power stability. It was able to bring the system output within 1.00% of the target power level in under 0.27 seconds. PSO consistently outperformed GA across different controllers in achieving faster convergence and better tuning results. Conversely, the FOPID controller, when optimized using GA, delivered the poorest performance, exhibiting a significant steady-state error of 6072.3W and noticeable overshoot and undershoot issues. To further improve operational efficiency, an ingenious switching mechanism was introduced. This algorithm dynamically selects the most cost-effective and demand-responsive energy source, enhancing the economic viability of the nano-grid. A case study was also conducted, demonstrating the system’s resilience: in the event of a fault in one of the generation units, the smart algorithm automatically rerouted power from alternative sources to maintain uninterrupted supply to the connected loads.