Combining the Wolf Pack Algorithm with the Curve-adaptive Grasshopper Optimization Algorithm

To address the issues of the Grasshopper Optimization Algorithm (GOA) falling into local optima and achieving low optimization precision, this paper proposes a hybrid algorithm called Combining the Wolf Pack Algorithm (WPA) with the Curve-adaptive Grasshopper Optimization Algorithm (WCGOA). Firstly, the population is initialized using Logistic mapping to ensure an optimal initial population, thereby enhancing population diversity. Secondly, the linear weights are replaced with curve-adaptive weights to improve search speed and precision. Thirdly, by incorporating the hierarchical hunting concept from the Wolf Pack Algorithm, individual awareness of grasshoppers is developed to enhance global search capabilities. Subsequently, Cauchy mutation is applied to the best individuals to strengthen their ability to escape local optima. Finally, GOA is compared with four other classical algorithms as benchmarks for WCGOA. Statistical analysis and Wilcoxon rank-sum tests conducted on 11 commonly used benchmark functions demonstrate that WCGOA significantly outperforms other algorithms in terms of convergence precision, convergence speed, stability, and optimization success rate.