Multi-strategy enhanced Grey Wolf Optimizer for global optimization and real world problems

The Grey Wolf Optimizer (GWO) is one of the more successful swarm-based intelligent algorithms in recent years, but the shortcomings of the Grey Wolf Optimizer are revealed as the problems handled become progressively more complex. For this purpose, this paper presents a new variant of GWO and names its Hybrid Contact List Subpopulation Mixed Evolution Grey Wolf Optimizer (CSELGWO). In the paper first introduces the Contact List Mechanism (CLM) to obtain high quality local optimal information in the search space. This is followed by the Hybrid Contact List Subpopulation Generation (HCSG) mechanism, which utilizes the information in the Contact List to assist in the updating of the Subpopulation and interacts with the main population through Subpopulation Mixed Evolution (SME) to interact with the main population, thus significantly improving population diversity and convergence accuracy. In addition, the proposed Levy Flight with archives and Activation Mechanism (LFAA) can moving away from local optimality by reasonable judgment. We evaluated it using 66 test functions and showed excellent convergence speed, stability and accuracy. Additionally, when compared with the top-performing algorithm from the CEC2020 Real World Competition, CSELGWO demonstrates effective solutions to real-world problems. Finally, we compared LSHADE_cnEpSin with LSHADE_SPACMA. Although CSELGWO does not outperform these LSHADE variants in terms of convergence accuracy and standard deviation obtained, it shows excellent performance on certain types of functions, indicating excellent potential. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.