Multi-swarm improved Grey Wolf Optimizer with double adaptive weights and dimension learning for global optimization problems

Grey Wolf Optimizer (GWO) is a swarm intelligent optimization algorithm that simulates the leadership and social behavior of grey wolves to prey. GWO is extensive used in various fields since it has the advantage of being simple and easy to implement. However, when solving complex optimization problems, GWO has insufficient population diversity and unbalanced exploration and exploitation. To overcome these shortcomings of GWO, a multi-swarm improved grey wolf optimizer (MIGWO) is proposed. In MIGWO, firstly, chaotic grouping mechanism is utilized to improve population diversity and dynamic regrouping mechanism to further improve population diversity and balance exploration and exploitation. Secondly, double adaptive weights and dimension learning are utilized to ameliorate the hunting behavior of grey wolves, which can improve search performance. The MIGWO is verified on 53 test problems. The test results show that it is better to other metaheuristic algorithms and GWO variants in solving the optimal solution and convergence accuracy. In addition, MIGWO is utilized to solve 4 engineering design problems.(c) 2022 International Association for Mathematics and Computers in Simulation (IMACS). Published by Elsevier B.V. All rights reserved.

Automated summary

Grey Wolf Optimizer (GWO) is a swarm intelligent optimization algorithm that simulates the leadership and social behavior of grey wolves to prey. GWO is extensively used in various fields due to its simplicity and ease of implementation. However, it has shortcomings in terms of population diversity and exploration-exploitation balance. To overcome these issues, a multi-swarm improved grey wolf optimizer (MIGWO) is proposed. MIGWO utilizes mechanisms like chaotic grouping, dynamic regrouping, double adaptive weights, and dimension learning to enhance the search performance. Experimental results demonstrate the superiority of MIGWO over other metaheuristic algorithms and GWO variants.