A Novel Hybrid Metaheuristic Based on Augmented Grey Wolf Optimizer and Cuckoo Search for Global Optimization
Satyam Sharma, Ridhi Kapoor, Sanjeev Dhiman
Abstract
Grey Wolf Optimizer (GWO) is an intelligent metaheuristic approach which imitates the leadership hierarchy and cooperative hunting behavior of a group of Grey wolves(wolf-pack). An augmentation of GWO, named Augmented GWO (AGWO), was recently proposed which possesses greater exploration abilities. Nevertheless, in some cases, AGWO underperforms in the exploitation phase and stagnates at local optimum. The CS algorithm is a nature-inspired optimizing technique that mimics the unique nesting strategy of cuckoo birds and levy-flights. Both the algorithms possess powerful searching capabilities. In our research work, a novel hybrid metaheuristic, termed AGWOCS, is put forth, which combines the merits of both metaheuristics in order to attain global optimum effectively. The proposed algorithm amalgamates the exploring abilities of the AGWO with the exploiting abilities of the Cuckoo Search (CS). For the purpose of testing the proficiency of our proposed hybrid AGWOCS, twenty-three renowned benchmarking functions. It is compared with six other existing metaheuristics, including Standard GWO, Particle Swarm Optimization (PSO), Augmented-GWO (AGWO), Enhanced-GWO (EGWO), Hybrid GWO with CS (CS-GWO) and Hybrid PSO and GWO (GWOPSO). The simulation results indicate that AGWOCS surpasses other metaheuristics in terms of rapid convergence rates as well as avoiding local optimum stagnation.