Refining swarm behaviors with human-swarm interaction strategies: An improved monkey algorithm for multidimensional optimization problems

Abstract This study introduces human-swarm interaction (HSI) strategies to enhance bio-inspired swarm intelligence (SI) algorithms, addressing inherent limitations of the traditional monkey algorithm (MA) such as premature convergence and computational inefficiency in complex search spaces. We propose three HSI integration strategies involving intermittent, persistent, and parameter-setting interactions within the HSI to augment emergent behaviors and refine the MA’s intrinsic optimization mechanisms. Validation through seven benchmark functions (one unimodal and six multimodal) across seven dimensions demonstrates the HSI-MA’s ability to resolve complex, multidimensional optimization problems with statistically significant (p < 0.05) superior accuracy and stability compared to the original MA and four baseline SI algorithms, achieving 85% dominance in test cases while reducing iterations by an order of magnitude. Further evaluation on five engineering design problems reveals the HSI-MA outperforms 36 state-of-the-art optimizers in 70% of scenarios, confirming its enhanced precision and efficiency in practical applications. In contrast to conventional fusion-based approaches, the HSI framework preserves the original algorithm’s theoretical foundations while systematically integrating human intelligence to enhance structural adaptability and operational efficiency.