Abhishek Bajirao Katkar, H. T. Jadhav
Renewable energy integration brings uncertainty to power systems, rendering traditional Optimal Power Flow (OPF) methods insufficient for reliable and sustainable operation. This study creates a detailed Multi-objective Optimal Power Flow (MOOPF) framework that models renewable intermittency by probabilistically characterising wind and solar power with Weibull and lognormal distributions. A new hybrid gbest-guided Artificial Bee Colony – NSGA-II (gbestABC – NSGA-II) algorithm is proposed to tackle the nonlinear, non-convex, and constrained optimisation challenge. This hybrid approach combines the exploration power of the Artificial Bee Colony with the elite sorting and crowding mechanisms of NSGA-II, ensuring better convergence and diversity. A Diversity-Enhanced Tri-Stage Repair (DEST) strategy ensures feasibility under complex operational limits, while a modified entropy-based TOPSIS module identifies the Best Compromise Solution (BCS). The MOOPF framework is validated using IEEE 30 and 57-bus test systems with wind and solar generation units. Simulation results show that the hybrid algorithm converges faster, offers better Pareto front diversity, and improves feasibility compared to benchmark metaheuristics like SDCS, MG-JAYA, EBCM, and HS-ES. Quantitative evaluation with Generational Distance (GD), Spacing Parameter (SP), and Diversity Metric (DM), plus Wilcoxon signed-rank tests, confirms the statistical significance of improvements. Optimised dispatch solutions reduce total generation costs by 8.7%, emissions by 10.3%, and power losses by 7.9%, all while ensuring voltage stability. The proposed hybrid gbestABC – NSGA-II framework is a scalable and robust tool for multi-objective stochastic OPF with renewable integration. It enhances the field by merging adaptive swarm-evolutionary cooperation with dynamic constraint management, offering a practical route to carbon-neutral and resilient smart grids.