A New Variant Of Grey Wolf Optimizer For Global Optimization

Abstract

The Grey Wolf Optimizer is the meta-heuristics which is inspired by Grey Wolves (Canis Lupis) and is used in SI algorithms. The GWO and proposed algorithm mimics the leadership hierarchy and hunting method of grey wolves. The four levels of grey wolves are alpha, beta, delta, and omega and the three main steps of hunting are searching, encircling, and attacking prey. When the wolves change from its position to attack a prey then the position has to be updated. In the GWO algorithm, the vector location is updated by taking a simple average of best location in the pack. In proposed algorithm the location is updated by taking an average of the weighted sum of the best locations. The main algorithm is compared with other SI algorithm and tested over the different benchmark functions. The dimensions of the problems are 10, 30 and 50 for comparing the proposed algorithm over the basic state of-the-art. Simulation results supports the better performance of proposed algorithm.