Power System Optimization Using Hybrid Heuristic Search Algorithm

Abstract

This research introduces a hybrid optimization algorithm integrating Genetic Algorithm (GA) and Greedy Algorithm for microgrid load prediction and operational control. The approach leverages GA's global exploration capabilities to identify potential configurations and applies the Greedy Algorithm for local refinement, addressing challenges such as slow convergence and suboptimal solutions. Real-world load data is utilized to evaluate the method's performance across forecasting accuracy, operational cost reduction, system reliability, and computational efficiency. Metrics like Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) validate the effectiveness of the proposed hybrid approach. Results demonstrate that the algorithm significantly improves load forecasting precision and microgrid control performance compared to traditional methods. By optimizing energy management and incorporating renewable energy sources, this method enhances sustainability, reliability, and efficiency, establishing itself as a viable solution for modern microgrid systems. This study contributes to advancing hybrid heuristic algorithms for power system optimization.