An enhanced dung beetle optimization algorithm based-on multi-strategies for solving global optimization problems

The Dung Beetle Optimization (DBO) algorithm exhibits rapid convergence and robust search capabilities, yet its performance is constrained by excessive reliance on global best and worst solutions. To resolve these weaknesses, this paper introduces an enhanced DBO that incorporates multiple strategies, named DCWDBO. The dynamic opposition-based learning mechanism improves the quality of the initial population. Horizontal and vertical crossover strategies are incorporated to strengthen search capabilities. To preserve high population diversity throughout iterations, the original boundary-control mechanism is replaced with rules from the Wave Search Algorithm. To evaluate DCWDBO’s effectiveness, it was compared with PSO, SCA, SCSO, and standard DBO using benchmark functions from CEC 2017, 2020, and 2022. Results indicate that DCWDBO achieves reliable performance, demonstrating robust global exploration, stable convergence, and superior large-scale optimization capability.