A Hybrid Spherical Search and Sine Cosine Algorithm
Kaiyu Wang, Sichen Tao, Zhiming Zhang, Yuki Todo, Shangce Gao · 2020
Recently, a metaheuristic algorithm called spherical search (SS) has been proposed. The main operations of SS are the calculation of spherical boundary and generation of a new trial solution on the surface of the spherical boundary. Although SS has shown superior performance over many metaheuristic algorithms. It is still easily to fall into local optimum and has poor convergence accuracy. In this paper, we for the first time propose a hybrid spherical search and sine cosine algorithm aimed to leverage the strengths of two different search mechanisms with a co-evolutionary manner. We use the opposite solution to avoid the blind learning from the current optimal solution, to effectively alleviate the immature convergence of the algorithm. Experiments based on 30 benchmark functions of IEEE CEC2017 are conduced and results demonstrate that the proposed algorithm outperforms other states-of-the-art algorithms.