Share
Export Citation
Implementation of Ant Colony Optimization and Particle Swarm Optimization for Power Plant Operation Optimization
Harris Hamma A.
Proceedings 2025 4th International Conference on Electronics Representation and Algorithm Artificial Intelligence Creating Tomorrow S World Today Icera 2025
Abstract
The increasing complexity of modern power systems necessitates advanced optimization techniques for economic dispatch (ED) to minimize fuel costs while ensuring system reliability. Traditional mathematical approaches often struggle with the nonlinear, multi-modal, and constrained nature of real-world ED problems, leading to the adoption of heuristic and metaheuristic algorithms. Among these, Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) have demonstrated superior performance compared to conventional methods. However, PSO suffers from premature convergence, while ACO exhibits slow convergence due to its pheromone update mechanism. To address these limitations, this study proposes a methodological integration of PSO and ACO, leveraging PSO's global exploration capability and ACO's local refinement mechanism to improve solution accuracy and computational efficiency. The proposed approach was tested on IEEE standard test systems (5-bus, 26-bus, and 30-bus) as well as real-world data from PLTD Selayar, with performance evaluations based on total fuel cost, convergence speed, and computational efficiency. Experimental results indicate that the combined ACO-PSO approach outperforms standalone PSO and ACO, achieving up to 8.7% lower fuel costs than PSO and reducing computational time by 32.5% compared to ACO in large-scale systems. Furthermore, real-world validation confirmed that the ACO-PSO combination is effective for economic dispatch planning, providing a more costefficient and reliable power generation strategy. Although the proposed method demonstrates clear advantages, parameter sensitivity and adaptation to dynamic load variations remain challenges requiring further refinement. Future research should explore self-adaptive parameter tuning, real-time optimization frameworks, multi-objective environmental dispatch, and integration with AI-driven forecasting models. This study contributes to the advancement of intelligent power system optimization by presenting an enhanced combination of PSO and ACO for improving economic dispatch performance in modern power grids.