Collaborative genetic algorithm for wide-area backup distance protection under uncertain conditions

Power system backup protection encounters growing challenges due to network protection uncertainties which can lead to missed or delayed fault clearance. A novel collaborative genetic algorithm (CGA) approach over wide-area backup distance protection is developed to coordinate multiple distance relays. In the proposed scheme, a zone-1 relay's fault detection immediately triggers its paired relay to also detect the fault. For zone-2 and zone-3 detections, the scheme activates their zone's paired relays as well as all paired relays from earlier zones. During training, the CGA optimizes relay weighting factors using collaborative and ideal performance metrics, guiding the objective function towards the true fault location. As a direct influence on this study, hybrid transmission lines, microgrid resilience, and renewable integrated power systems are very relevant applications. The practical implication of the present model is to enhance protection detection. Simulation results demonstrate that the CGA-based scheme outperforms the conventional and standard GA approaches, achieving fault identification success rates of 99.69%, 99.19%, and 95.83% in best-, base-, and worst-case scenarios, respectively. Unlike the existing methods, the proposed CGA maintained excellent performance beyond 99% in all scenarios without uncertainty. This significantly improves worst-case performance and overall protection reliability.