Abstract: The individuals in the grey wolf group tend to approach to the area where the leadership grey wolf is located because of the reduction of attenuation factor in the middle and late stage of the optimization process for the standard grey wolf optimization algorithm, which results in poor global optimization ability and reduces the optimization accuracy of the algorithm.Aiming at this problem, an improved grey wolf optimization algorithm (IGWO) is proposed.Firstly, by analyzing the influence of the decay factor, an adjustable segmentation decay factor is presented to focus on proper balance between exploration ability and exploitation ability.The appropriate parameter is searched to achieve more precise optimization.And it ensures that the algorithm has a certain global search ability in the middle and later stages of the optimization process.The numerical simulation experiments show that increasing the exploration ratio is beneficial to improve the convergence accuracy of the algorithm.Meanwhile, leadership wolves are selected by probability to carry out levy flight operation or random walk operation respectively in the process of optimization.The global optimization ability of the algorithm is improved by the levy flight with the feature of short-range flight search and occasional long-distance walk.While the local optimization ability is improved by random walk with relatively centralized search range.Simulation experiments are conducted on optimization problems.The results show that compared with other algorithms, the improved algorithm has great performance in optimization precision, algorithm stability and convergence speed.

Key words: Decay factor, Grey wolf optimization algorithm, Levy flight, Random walk