Adaptive Bitterling Fish Optimization with Evolutionary Game Theory: For Cross-Regional Emergency Repair Path Planning

Modern energy internets and large-scale industrial systems are becoming increasingly complex. Consequently, the rapid response capability of energy infrastructure during sudden failures has become a core element to ensure the stable operation of the social economy. Emergency repair path planning (ERPP) is a complex nonlinear combinatorial optimization problem. It is characterized by dynamic uncertainties, such as fluctuating task durations and variable traffic accessibility. This paper proposes a cross-regional emergency repair path planning (CR-ERPP) optimization model considering dynamic path conditions. The model takes into account jurisdiction ownership, cross-regional dispatch costs, path weights (congestion coefficient, grade coefficient, quality coefficient) and accident risk levels. The primary objective of this model is to minimize the total repair cost. Furthermore, an Adaptive Bitterling Fish Optimization with Evolutionary Game Theory (ABFO-EGT) is developed. It introduces adaptive mechanisms, evolutionary game theory, and a symmetric mutation strategy. These enhancements are designed to overcome the inherent limitations of traditional swarm intelligence algorithms, namely unbalanced search behavior and premature convergence to local optima. Performance analysis demonstrates that the ABFO-EGT algorithm exhibits superior convergence stability and global search capability. Case study results show that the proposed method significantly reduces the total repair cost. Specifically, the cost is reduced by 33.2% compared to manual decision-making, 27.6% compared to the GWO algorithm, and 9.1% compared to both the ACO and PSO algorithms. This study provides an efficient and reliable decision support tool for emergency management of large-scale energy systems.