A dual-state epsilon driven hybrid Gbest Artificial Bee Colony–NSGA-II framework for stochastic renewable multi-objective optimal power flow

Abstract Integrating renewable energy sources adds uncertainty, nonlinearity, and complexity to multi-objective optimal power flow (MOOPF) problems, requiring strong and efficient optimization frameworks. This paper presents an improved hybrid Gbest-guided Artificial Bee Colony and NSGA-II (GbestABC–NSGA-II) algorithm with a dual-state epsilon-based constraint handling strategy to enhance feasibility, convergence, and solution diversity. The framework combines global-best guided exploitation with elitist non-dominated sorting for a balanced exploration-exploitation trade-off. The dual-state epsilon mechanism manages constraint violations, while epsilon-dominance archiving improves Pareto front convergence and diversity. Realistic thermal generator modeling is enhanced by using multi-fuel cost functions, valve-point loading effects, and prohibited operating zones for better practical applicability. Renewable uncertainty is modeled with Weibull-distributed wind speed and lognormal solar irradiance for precise stochastic representation. A BWM–TOPSIS decision-making approach is used to find the best compromise solution. The proposed method’s effectiveness is validated on modified IEEE 30-bus and 57-bus systems across various operational scenarios, including renewable integration and security constraints. The results show notable performance gains, with up to a 7% decrease in generation costs, around 24–26% lower emissions, and a significant drop in transmission losses compared to traditional methods. The statistical evaluation using IGD, HV, PDI, and Wilcoxon rank-sum test shows the proposed algorithm’s superior convergence, diversity preservation, and robustness. The findings confirm that the hybrid framework efficiently and reliably solves complex renewable-integrated MOOPF problems, improving techno-economic and environmental performance.