Multi-population-based adaptive sine cosine algorithm with modified mutualism strategy for global optimization

The sine cosine algorithm (SCA) is a population-based metaheuristic strategy that has been demon-strated competitive performance and has received significant attention from scientists in various fields. Regardless, like other population-based techniques, SCA also has a tendency to get stuck in adjacent optima and uneven exploitation. Given the shortcomings of SCA, a new modified SCA variation, MAMSCA, with a balanced global and local search, is presented in this work. The new method divides the population into two equal halves for updating using a sine or cosine strategy. To provide additional variation to the population, a modified mutualism phase is adopted. To increase convergence speed and accuracy of the solution, a one-to-one mapping between the individuals of each half is maintained. The new algorithm is compared with state-of-the-art algorithms and modified algorithms using classical benchmark functions and IEEE CEC 2019 functions. The method's real-world relevance is demonstrated by tackling five engineering design challenges. The comparison of numerical results and analysis of performance measures of the algorithm from the statistical aspect, time complexity, and speed of generating the solution demonstrates the considerable improvement of the proposed algorithm to solve real-world challenges. (C) 2022 Elsevier B.V. All rights reserved.

Automated summary

This paper presents a modified sine cosine algorithm (MAMSCA) that addresses the shortcomings of the original algorithm by balancing global and local search and introducing additional variation to the population. The proposed algorithm demonstrates significant improvement in solving real-world challenges.