Multi-Constrained Multi-Objective Modeling and Algorithm for Hierarchical Protection Site Selection in Distribution Networks

To address the lack of unified and quantitative criteria for hierarchical protection site selection in distribution networks, this paper proposes a multi-objective optimization framework for protection placement planning. The proposed method explicitly quantifies the positional relationships among protection devices, the allowable number of protection configurations, and the installation constraints of mainline protection. Based on these factors, a multi-objective optimization model is established with the objectives of minimizing total protection operating time, load loss, and investment cost. An improved bare-bone multi-objective particle swarm optimization (BB-MOPSO) algorithm is developed to solve the discrete and constrained site selection problem, in which dynamic constraint dominance and crowding-distance-based selection are introduced to enhance convergence and solution diversity. To reduce the complexity of multi-objective decision making, K-means clustering combined with min–max normalization is employed to extract representative protection schemes from the Pareto-optimal solution set. Simulation on the IEEE 33-node distribution network demonstrates that the proposed approach reduces protection operating time, load loss, and equipment investment by up to 12.05%, 8.12% and 18.17%, respectively, thereby improving the rapidity, selectivity, and economic performance of distribution network protection.