Improving Arithmetic Optimization Algorithm Through Modified Opposition-based Learning Mechanism

This study aims to present a novel hybrid metaheuristic algorithm through improving the performance of the arithmetic optimization algorithm (AOA). A modified version of opposition-based learning mechanism (mOBL) has been used to provide the improvement. The greater performance of the improved version of the arithmetic optimization algorithm (mOBL-AOA) has been demonstrated through statistical and non-parametric tests by using benchmark functions of Schwefel 2.22, Rosenbrock, Step, Schwefel, Ackley and Penalized. The results were demonstrated comparatively by using sine cosine, Levy flight distribution and the original arithmetic optimization algorithms. The performed comparative analyses have confirmed the highly competitive performance of the mOBL-AOA algorithm in terms of tackling with the the optimization problems.