Sine Cosine Algorithm with Centroid Opposition-Based Computation

Abstract

Sine cosine algorithm (SCA), a population-based optimization algorithm, is recently developed to solve optimization problems. In SCA, mathematical functions sine and cosine are utilized to fluctuate the candidate solutions towards or outwards the best solution. SCA gets trapped in local optima and suffers from premature convergence for some problems due to a lack of exploration of the search space. In this paper, SCA is improved by incorporating an opposition-based learning (OBL) scheme called centroid opposition-based computing (COBC) in it to enhance its exploration ability. The proposed algorithm is termed as COBSCA in this paper. It is applied to solve 28 CEC2013 benchmark problems. The results of COBSCA are compared with SCA and opposition-based SCA (OBSCA). The experimental results demonstrate that the COBSCA statistically outperforms others in solving most of the problems in the CEC2013 benchmark suite.