Enhanced Social Group Optimization Algorithm for Solving Optimization Problems

Abstract

In the last decades, the field of global optimization has experienced significant growth, leading to the development of various deterministic and stochastic algorithms designed to tackle a wide range of optimization problems. One notable member of this family is the Social Group Optimization (SGO) algorithm. The Improving Phase and the Acquiring Phase are its two main fundamental phases. The two upgraded versions of SGO with a modified improvement phase are Enhanced Social Group Optimization (ESGO) and Enhanced Modified Social Group Optimization (EMSGO). The key enhancement in these variants focuses on honing, refining skills and abilities to achieve greater effectiveness. To assess the performance of ESGO and EMSGO, an extensive comparative analysis is conducted, involving twelve algorithms, including recently introduced, potent metaheuristic methods. Since both ESGO and EMSGO are modified algorithms, a comparison is conducted between these two algorithms and six recently introduced improved/hybrid algorithms. Subsequently, twenty-six real-world design problems from the mechanical and chemical engineering areas are addressed by applying both modified methods. The simulation results leave no doubt about the capability of ESGO and EMSGO to consistently achieve optimal solutions. Their robust performance, both in comparative evaluations and real-world applications, underscores their potential in solving challenging optimization tasks.