Hybrid Optimization Algorithm Based on Arithmetic and Golden Sine Algorithms for Constrained Engineering Problem

Arithmetic Optimization Algorithm (AOA) is a newly proposed meta-heuristics (MAs) inspired by the major arithmetic operators. Although it provides satisfactory solutions to several real-world problems, when facing complex problems, AOA also easily suck into local optimal solution and slow convergence speed. Golden Sine Algorithm (Gold-SA) is another MAs and has an acceptable local search capability with fewer parameters. Considering these situations, this paper we propose a novel hybrid optimization algorithm (HAGA) that integrated both AOA and Gold-SA to achieve superior optimization ability. In HAGA, the search agents are divided into two consistent subgroups and optimized by AOA and Gold-SA, respectively. By this operator, we can exchange useful information and utilize their advantages to find the optimal or near-optimal regions throughout the iterative process. In addition, Lens Opposition- Based Learning (LOBL) is employed to accelerate the convergence speed. We test the proposed method using thirteen benchmark functions and a constrained engineering problem. The results prove the significant superiority of HAGA to those of four counterpart methods in terms of optimization accuracy and convergence speed.