Sooty Shearwater Migration Optimizer: A Novel Metaheuristic Algorithm for Engineering Optimization and Photovoltaic Parameter Identification

This paper develops a new metaheuristic optimizer, named the Sooty Shearwater Migration Optimizer (SSMO), by drawing inspiration from the long-distance migration patterns of sooty shearwaters. The proposed SSMO integrates elite guidance, wind-driven migration, molting memory guidance, and adaptive foraging to coordinate global exploration with local exploitation. Its optimization performance is assessed in the CEC2017 benchmark suite under 30D, 50D, and 100D settings and in the CEC2022 benchmark suite under 10D and 20D settings. Compared with 11 representative optimization algorithms, SSMO achieves the lowest average Friedman ranks in all five-dimensional settings. The Wilcoxon rank-sum results also generally favor SSMO. Additional comparisons with four differential evolution variants further demonstrate competitive performance under the tested configurations. Its practical applicability is further assessed using five engineering design problems, on which it obtains the lowest mean fitness values under the adopted evaluation settings. In addition, four photovoltaic parameter identification cases are used to evaluate its effectiveness in practical parameter estimation. Compared with the selected general-purpose algorithms, SSMO obtains the lowest error values for the considered evaluation metrics. Further comparison with four specialized photovoltaic parameter identification algorithms shows that SSMO remains competitive in most test cases. These findings suggest that SSMO is a promising optimizer for numerical benchmarks and practical engineering applications.