Guaranteed Convergence Sine Cosine Algorithm
Yunfeng Zou, Juan Zhao, Zheng-Ming Gao · Proceedings of the 2021 5th International Conference on Electronic Information Technology and Computer Engineering · 2021
We are still under the way in finding new optimization algorithms including their improvements to conquer the complicated problems and find their solutions. Considering the updating equation in the sine cosine (SC) algorithm, the positions in the next iteration would be an direct addition operation of the current positions and their random weighted, sine or cosine functionalized, absolute distance to the position of the global best candidate, which would slow down the approaching rate when the individuals were near the global optima, or decrease the capability when the individuals were trapped or faraway from the global optima. To conquer this shortcomings, the guaranteed convergence method is introduced to increase the SC algorithm in this papaer. Simulation experiments are carried out and qualitative, intensification, diversification analysis are made. Results verify the capability in superiority of the improved algorithm in optimizing traditional symmetric benchmark functions, either unimodal or multimodal. However, both of them lack the capability in optimzing non-symmetric benchmark functions, although the guaranteed convergence SC algorithm still performed overall better.