Abstract: To improve relay protection performance in distribution networks with high-penetration distributed generation (DG), this study proposes a multi-scenario optimization method that balances selectivity and speed. A DG output-fault current coupling model is built, and 5000 Monte Carlo scenarios are generated. A multi-objective function is formulated using selectivity overlap and clearing time, and optimized via an improved NSGA-II algorithm. Results show a 32% reduction in average operation time and a drop in misoperation rate from 5.2% to 1.1%. The method demonstrates strong adaptability and provides a practical solution for protection setting in DG-rich distribution networks.