An Adaptive Fitness-Guided Starfish Optimization Framework for Optimal Power Flow Operation

Optimal Power Flow Operation (OPFO) is a large-scale, nonlinear, and highly constrained optimization problem that plays a central role in achieving economical, reliable, and environmentally sustainable power system operation. Despite the widespread use of metaheuristic algorithms for OPFO, many methods primarily depend on global-best updates or complex hybrid operators, leading to issues like premature convergence and diminished population diversity. Furthermore, recent literature tends to focus on numerical improvements without sufficiently addressing the underlying interaction structures that ensure stability in convergence. To address these limitations, this paper proposes an Improved Starfish Optimization (ISFO) algorithm incorporating a hybrid fitness-aware population-based search mechanism for solving OPFO problems involving the simultaneous regulation of synchronous generator outputs, on-load tap-changing transformer ratios, and reactive power compensation devices. The proposed method introduces an adaptive Fitness-Aware Collective (FAC) interaction strategy that systematically models pairwise fitness relationships to guide attraction toward superior solutions and repulsion from inferior ones, thereby strengthening exploitation while preserving diversity through controlled stochastic peer-based perturbations. A dual-mode search framework further balances global exploration and local intensification without introducing additional control parameters, enhancing robustness and scalability. The OPFO problem is formulated as a constrained nonlinear optimization model, where equality constraints enforce power flow balance equations and inequality constraints represent operational limits of generators, transformers, voltages, and transmission lines. The proposed ISFO is validated on the IEEE 57-bus power system under three operating scenarios: fuel cost minimization, transmission loss minimization, and emission minimization. Comparative results demonstrate consistent superiority over the standard Starfish Optimization Algorithm (SFOA). In cost minimization, ISFO reduces the total generation cost from 41, 697. 85 /h to 41, 669. 34 /h while simultaneously decreasing real power losses by 5. 22%. Under loss minimization, ISFO achieves a minimum transmission loss of 10. 77 MW, corresponding to a 9. 23% reduction relative to SFOA, with improved convergence stability. For emission minimization, ISFO attains the lowest emission level of 1. 474 ton/h, representing a 6. 65% reduction compared to SFOA, alongside an additional 5. 67% reduction in system losses. Statistical evaluations based on 30 independent runs further confirm the robustness and reliability of the proposed approach, demonstrating reduced variance, narrower confidence intervals, and statistically significant improvements across all investigated objectives.