Research on adaptive coati optimization algorithm based on chaos and perturbation strategies

The coati optimization algorithm (COA) is a bionic optimization algorithm that imitates natural phenomena and has good performance, but it suffers from the limitations of slow convergence and low accuracy of the optimal solution. Therefore, in this paper, an improved meta-heuristic algorithm called the adaptive sine coati optimization algorithm (ASCOA) is proposed. The proposed algorithm introduces a chaotic mechanism to optimize the initialized population. Cauchy perturbation and Levy flight are added in the exploitation phase to dynamically adjust the position updating method with probability to avoid falling into a locally optimal solution. And adaptive weights are added as a way to comprehensively improve the overall algorithmic optimality-seeking ability. The performance of ASCOA is tested in different dimensions using 13 sets of test functions, comparing ASCOA with well-known metaheuristic algorithms through numerical results as well as convergence curves. ASCOA is also applied to engineering optimization. Simulation results show that the ASCOA can balance exploration and exploitation and improve the convergence speed and numerical accuracy. The Wilcoxon rank sum test shows that the results obtained in the paper are statistically significant. © 2024, Beijing University of Posts and Telecommunications. All rights reserved.