Adaptive role division optimizer: An efficient socio-inspired meta-heuristic algorithm with dynamic proportion adjustment

Meta-heuristic algorithms are established as effective tools for solving complex optimization problems across various disciplines. However, when confronted with high-dimensional and non-convex search spaces, existing algorithms often face challenges such as premature convergence and insufficient global exploration capabilities. To address these challenges, we propose an Adaptive Role Division Optimizer (ARDO) inspired by social division of labor. ARDO models a population composed of three dynamically assigned roles, pioneers for global exploration, coordinators for local exploration, and executors for local exploitation. Our ARDO then adaptively adjusts role distribution and population structure through a population convergence factor and a social mobility mechanism, which enables an autonomous balance between exploration and exploitation. This flexible design mimics the adaptive regulation of natural populations, allowing effective self-adjustment based on ongoing optimization state and environmental feedback. We evaluate the ARDO’s performance against ten prominent meta-heuristic algorithms on the CEC 2017 and CEC 2022 benchmark functions. Experimental results show that ARDO achieves consistently superior solution accuracy and robustness. Its practical effectiveness is further demonstrated through five practical optimization cases. The source code will be publicly available at https://github.com/windchantofjade/ARDO-optimizer.