An Efficient and Fast Hybrid GWO-JAYA Algorithm for Design Optimization

Abstract

Metaheuristic algorithms (MHAs) are widely used in engineering applications in view of their global optimization capability. Researchers continuously develop new MHAs trying to improve the computational efficiency of optimization search. However, most of the newly proposed algorithms rapidly lost their attractiveness right after their release. In the present study, two classical and powerful MHAs, namely the grey wolf optimizer (GWO) and the JAYA algorithm, which still attract the attention of optimization experts, were combined into a new hybrid algorithm called FHGWJA (Fast Hybrid Grey Wolf JAYA). FHGWJA utilized elitist strategies and repairing schemes to generate high-quality new trial solutions that may always improve the current best record or at least the old population. The proposed FHGWJA algorithm was successfully tested in seven engineering optimization problems formulated in the fields of robotics, hydraulics, and mechanical and civil engineering. Design examples included up to 29 optimization variables and 1200 nonlinear constraints. The optimization results proved that FHGWJA always was superior or very competitive with the other state-of-the-art MHAs including other GWO and JAYA variants. In fact, FHGWJA always converged to the global optimum and very often achieved 0 or nearly 0 standard deviation, with all optimization runs practically converging to the target design. Furthermore, FHGWJA always ranked 1st or 2nd in terms of average computational speed, and its fastest optimization runs were better or highly competitive with those of the best MHA taken for comparison.