Research On The Improvement Of Grey Wolf Optimization Algorithm

Grey wolf optimization algorithm(GWO)is a meta-heuristic optimization technique developed by imitating grey wolf's hierarchy and hunting methods.It has been successfully applied in path planning,cluster analysis,feature selection,power dispatching,etc.However,GWO also has the disadvantages of poor populationdiversity,slow convergence in later period and easy to fall into local optimization.Aiming at the shortcoming that the algorithm is easy to fall into local optimality,this paper proposes two new individual update mechanisms,and combines them with GWO to propose two new improved algorithms to make the algorithm jump out of local optimality by balancing the exploration and exploitation capabilities.The main research work is as follows:(1)Proposing an adaptive leadership grey wolf optimization algorithm(AL-GWO):In AL-GWO,an individual renewal mechanism based on the number of leaders is designed,and a parameter adaptive adjustment of the number of leaders is used to balance the exploration and exploitation capabilities.Experiments on 20 benchmark functions show that AL-GWO has better performance than other algorithms.(2)Proposing a grey wolf optimization algorithmbased on differential evolution(DE-GWO):Due to the original GWO has strong exploitation capability,this algorithm introduces the DE with strong exploration ability to update individual position.This not only expands the search scope of GWO but also achieves the balance between exploration and exploitation,and finally improves the defects of GWO.By applying DE-GWO to 20 benchmark functions,the results show that DE-GWO is superior to other algorithms in solving accuracy and convergence.