APPLICATIONS OF COMPUTATIONAL INTELLIGENCE TECHNIQUES FOR SOLVING THE OPTIMAL POWER FLOW PROBLEM IN MODERN POWER SYSTEMS

Abstract

The growing demand for electricity, rising greenhouse gas emissions, and the global emphasis on sustainability have led to the integration of intelligent optimization techniques in modern power systems.to ensure efficient, stable, and sustainable gird operations. This thesis addresses the Optimal Power Flow (OPF) problem, focusing on single-objective and multi- objective formulations to optimize economic, technical, and environmental parameters in modern power systems. Key objectives include fuel cost minimization, power loss reduction, voltage stability improvement, voltage deviation minimization, and emission reduction. The OPF problem is inherently complex, characterized by its nonlinear, nonconvex, and high-dimensional nature. The nonlinear, high-dimensional nature of the OPF problem is tackled using state-of-the-art metaheuristic algorithms, including the Learning-based Sine Cosine Algorithm (L-SCA), Hybrid Rao-2 Sine Cosine Algorithm (HRSCA), Coot Optimization Algorithm (COOT), and Electric Eel Foraging Optimizer (EEFO). Extensive testing has been conducted on standard IEEE networks, including the IEEE 30-bus, 57-bus, 118-bus systems, and the Algerian 59-bus network, to validate the scalability and robustness of these algorithms under varying operational scenarios. Single-objective and multi-objective formulations are analyzed to optimize control variables such as generator outputs, bus voltages, transformer tap settings, and reactive power compensation. Additionally, the integration of Distributed Generation (DG) units as constant power sources is investigated to assess the impact of renewable energy integration on system performance. The findings highlight significant improvements in system efficiency, reduced operational costs, enhanced stability, and reduced environmental impact. The proposed methodologies demonstrate rapid convergence, high-quality solutions, and computational efficiency, showcasing their applicability to real-world power systems. By addressing critical challenges such as fuel cost minimization, handling load growth scenarios, voltage collapse prevention, and emission reduction, this work contributes significantly to the development of sustainable and reliable energy systems. Future studies could explore the integration of advanced hybrid optimization techniques and real-time dynamic control systems to further enhance the efficiency and scalability of the proposed methodologies, particularly for large-scale, decentralized power systems with renewable energy integration.