Complementary Differential Evolution-based Whale Optimization Algorithm for Function Optimization

摘要

The whale optimization algorithm (WOA) is a metaheuristic search algorithm for solving the problem of function optimization. However, in the later stage of iterations, WOA suffers from premature convergence because the search agents are attracted by the elite vector. In this paper, a hybrid WOA based on complementary differential evolution, called CDEWOA, is proposed. First, a novel uniform initialization strategy is employed to enhance the diversity of initial population. Second, the differential evolution with a complementary mutation operator is embedded in the WOA to improve search accuracy and speed. Third, the introduction of a local peak avoidance strategy enables CDEWOA to jump out local optimum. Finally, the proposed CDEWOA is tested with 14 mathematical optimization problems. The test results illustrate that CDEWOA has better performance than IWOA, WOA, CDE, DE, and PSO in terms of convergence speed and convergence accuracy.