The rapid integration of renewable energy resources and electric vehicles (EVs) has significantly increased the operational complexity of modern power systems, particularly under uncertain generation and load conditions. In this context, the combined heat and power economic dispatch (CHPED) coupled with optimal power flow (OPF) emerges as a highly non-linear, large-scale, and multi-objective optimization problem that simultaneously seeks economic efficiency, environmental sustainability, and secure system operation. Although recent studies have demonstrated the effectiveness of advanced metaheuristics such as the quasi-oppositional sine cosine algorithm (QOSCA) in addressing this challenge, their fixed exploration–exploitation mechanisms and limited adaptability to dynamic uncertainty landscapes restrict further performance improvement.
To overcome these limitations, this paper proposes a novel hybrid optimization framework that integrates an Artificial Bee Colony algorithm enhanced by Online Fitness Landscape Analysis with NSGA-II (OFLA-ABC–NSGA-II). The proposed approach continuously monitors landscape characteristics such as modality, ruggedness, and basin transitions during the search process, enabling the optimizer to dynamically regulate global exploration and local exploitation. The ABC component exploits this landscape feedback to guide search behavior adaptively, while NSGA-II ensures effective Pareto dominance sorting and diversity preservation for multi-objective optimization.
The proposed framework is applied to uncertainty-aware CHPED-based OPF problems on the IEEE-57 and IEEE-118 bus systems with a high penetration of wind, solar, hydro, and EV resources. Comprehensive simulations under multiple operating scenarios demonstrate that OFLA-ABC–NSGA-II consistently outperforms the previously reported QOSCA and other state-of-the-art optimizers in terms of generation cost reduction, emission minimization, active power loss, average voltage deviation, voltage stability enhancement, and computational efficiency. Statistical validation using ANOVA and robustness analysis further confirms the superiority and reliability of the proposed method.
The results establish OFLA-ABC–NSGA-II as a powerful and scalable optimization framework for next-generation uncertainty-driven smart grid operation.
