Akademik Veri Yönetim Sistemi
CLUSTER COMPUTING, cilt.29, sa.1, ss.1-51, 2025 (SCI-Expanded, Scopus)
Optimal power flow (OPF) is one of the most remarkable optimization problems in the operation and planning of modern power systems. The complexity of the problem and the significance of the design of the problem both rise when renewable energy sources (RESs) and multi-terminal direct current systems are incorporated into electrical networks. Solving the non-convex OPF problem in modern electrical power systems with this structure is an important research topic. The main way to find the optimum solution set for the OPF problem whose objective functions are in conflict is to design a multi-objective evolutionary algorithm (MOEA) that exhibits the most appropriate search behavior for the geometric structure of the problem’s objective, constraint, and decision spaces. Recently, dynamic switched crowding (DSC) has been introduced as a new archive handling method for problem-specific tuning of the exploration/exploitation behavior of MOEAs. The DSC approach was used in this work to build 10 different versions of the multi-objective manta ray foraging optimization (MOMRFO) algorithm. The DSC-MOMRFO method, which shows the best search behavior for the OPF problem’s requirements, was also examined. The results of the MOMRFO algorithm, designed using the DSC version that exhibits the most appropriate search behavior, were analyzed comparatively with the results of the most used multi-objective optimization algorithms in the literature. The comparison results were evaluated using the Hypervolume (HV) performance metric, Wilcoxon and Friedman statistical test methods. According to Wilcoxon test results, it is seen that the performance of the DSC-based MOMRFO algorithm produces better results between 62.5% and 100% in solving different versions of the multi-objective OPF problem compared to other algorithms.