Research on multi-AUV cooperative 3D path planning

In the context of complex underwater environments, the advancement of Autonomous Underwater Vehicle (AUV)^[1] technology has brought the path planning problem for multi-AUV cooperative search^[2] into focus as a research hotspot. Traditional algorithms often struggle to efficiently generate high-quality routes in such environments, primarily addressing two-dimensional path planning problems. This paper aims to address the efficiency and capability limitations of traditional single-AUV path planning methods under multi-constraint and dynamic environments. A novel approach is proposed, leveraging cooperative navigation mechanisms among multiple AUVs to achieve effective path planning in threedimensional underwater environments^[3]. The algorithm first constructs a model of the operational area and then employs a distributed^[4] approach to assign navigation regions to each AUV. Path optimization is performed based on specific constraints such as detection capabilities, navigation speed, and obstacle avoidance performance of individual AUVs. Furthermore, the algorithm integrates real-time updates using new environmental information and employs a Bayesian model based on Markov properties for posterior information updates, enabling online discrete path planning. Experimental results demonstrate that the proposed multi-cooperative algorithm achieves significant improvements in task execution time and route quality compared to traditional methods, better addressing the demands of complex underwater scenarios.