Marine predators social group optimization: a hybrid approach.

Numerous heuristic and metaheuristic algorithms inspired by nature have been developed so far. These algorithms have demonstrated their excellence in resolving challenging problems across a variety of fields. However, many optimization algorithms fail while solving a complex real-world problem efficiently in an independent way because these problems have a distinct search behavior as they contain numerous equality and inequality constraints of the linear, nonlinear, and non-convex types. In this study, two optimization algorithms: the Marine Predators Algorithm (MPA) and the Social Group Optimization (SGO) algorithm are hybridized in order to increase their efficiency and problem-solving capability to solve this type of problem. This proposed hybrid optimization algorithm is named as Marine Predators Social Group Optimization (MPSGO) algorithm. This hybrid algorithm combines the strengths of both techniques to enhance the efficiency and effectiveness of the optimization process. It is validated through thirty benchmark functions of CEC2014 and after that 26 real-world optimization problems test suites of CEC 2020 from the chemical and mechanical engineering domain have been solved. The validation results are compared with ten state-of-art optimization algorithm and real-world optimization problem result is compared with BiPop matrix adaptation evolution strategy (BP-ϵMAg-ES), enhanced Multi-Operator Differential Evolution (EnMODE), LSHADE for Constrained Optimization (COLSHADE), Cohort Intelligence self-adaptive penalty function (CI-SAPF), and Cohort Intelligence self-adaptive penalty function Colliding Bodies Optimization (CI-SAPF-CBO). As MPSGO is a hybrid algorithm, again we have compared its performance with the seven best modified and improved optimization algorithms. In terms of convergence quality and improving solution quality, the simulated results of MPSGO have been shown to be competitive with those of state-of-the-art, hybrid, improved, and advanced optimization algorithms, proving the validity as well as the feasibility of the MPSGO algorithm. [ABSTRACT FROM AUTHOR]