Multipopulation Grey Wolf Optimization for Cooperative Multi-AUV Path Planning in Complex Underwater Environments

This article presents an innovative approach, termed the improved multiobjective multiple populations grey wolf optimization algorithm (IMMP-GWO), to address the cooperative planning problem of multiple autonomous underwater vehicles (multi-AUVs) in complex underwater environments with multiple target points. The algorithm comprehensively considers factors, such as AUV maneuverability, energy consumption, seabed topography, and obstacle avoidance. A key innovation is the hybrid enhancement using the greedy algorithm and the Cauchy mutation operator, which significantly boosts the algorithm’s evolutionary capabilities. In addition, the convergence factor of the traditional GWO algorithm is improved through the incorporation of the cosine law, addressing its inherent convergence limitations. A position update strategy, based on the sine cosine algorithm and iterative correlation of dynamic and static elements, is also introduced to better balance exploration and accelerate convergence. To tackle the orientation problem during GWO boundary processing, a novel approach is proposed that generates new individuals based on the minimum distance optimal weights. Simulation results demonstrate the superior performance of the IMMP-GWO algorithm in terms of computational complexity, convergence speed, and stability, validated through ablation studies and sensitivity analyses of key parameters. Overall, IMMP-GWO fulfills multi-AUV planning requirements and outperforms existing methods in various performance metrics.