Improving the convergence velocity of the balancing composite motion optimisation algorithm

Abstract

The balancing composite motion optimisation (BCMO), is a metaheuristic algorithm. However, the strengths of BCMO lie in its ability to extract information from the current generation and balance local and global searches. As a result, the convergence velocity of BCMO is superior to other optimisation methods, with a parameter-free and simplified computational procedure. This study proposes an enhanced algorithm, BCMO_A, designed to further increase the convergence velocity of the original BCMO. The proposed enhancement involves introducing an individual selection operator for the new generation. Three candidate options for the selection process were explored, showing minor differences. The best individuals from the two available populations were selected before interaction. The study also emphasises the importance of the terminal condition, which is based on the best value and the mean of objective functions in the current population. Numerical experiments were conducted on 23 classical benchmark functions, and the results were analysed using official statistical tests. The findings demonstrate that the original BCMO has been significantly improved, achieving better convergence speed, with optimised values reached earlier.