A multi-strategy fusion dung beetle optimization algorithm

A recently developed swarm intelligence method that converges more quickly than some existing algorithms is called the dung beetle optimization algorithm (DBO). However, the algorithm still suffers from the disadvantage that the initial population is not uniform and easily falls into the global optimal solution. In this paper, a multistrategy fusion dung beetle optimization algorithm (MSFDBO) is proposed which is inspired by the sine search algorithm, whale optimization algorithm, and reverse learning algorithm. First, the initial population of dung beetles is initialized using the singer chaos function to obtain a more uniform initial population. In view of the rolling behavior of dung beetles, adaptive coefficients and sine search strategies are introduced to enhance the search range of the algorithm, and the amplitude of the sine function is changed through the adaptive coefficients to balance the local and global searches of the algorithm. In view of the foraging behavior, egg-laying behavior and stealing behavior of dung beetles, an adaptive spiral search strategy is introduced to improve the optimization ability of the algorithm in the solution space. For the optimal solution obtained in each iteration, a convex lens imaging reverse learning strategy is used to perturb the optimal solution to avoid the dung beetle population from falling into a local optimum during the iteration process. By comparing the benchmark test function and the Wilcoxon rank sum test, MSFDBO shows better convergence performance, optimization performance, and robust performance.