随着水下业务规模的不断扩大，基于AUV集群的水下搜索系统由于其高度的冗余性、协同性、健壮性成为水下信息科学的研究热点。然而，同比与单AUV系统，AUV集群系统面临来自系统信息协同与控制效率两方面的挑战，尤其体现在针对水下特殊搜索任务的路径规划效率上。本课题引入网络化AUV集群深海搜索系统的概念，利用软件定义网络技术解耦系统的控制平面和数据操作平面。在此基础上，针对AUV集群协同路径规划机制拟从如下三个方面展开研究:（1）研究基于软件定义网络架构的多场景信息交互管理方案，以保障AUV集群的信息交互和控制效率；（2）研究基于自适应迭代贝叶斯滤波理论的协同定位方法和稳定性优化模型，以提高AUV集群定位的准确性；（3）研究抑制时间衰减率的改进GBNN模型以提高AUV集群路径规划的实时性。最终实现高效、透明、智能的面向AUV集群网络搜索系统的协同路径规划机制。

With the continuous expansion of the scale of underwater applications, the underwater searching system based on AUV cluster has become a research hotspot in underwater information science, due to its high degree of redundancy, coordination, and robustness. However, compared with single AUV-based system, the AUV cluster-based system confronts the challenges from the aspects of both system information coordination and control efficiency, especially in the path planning efficiency for some particular underwater exploration tasks. This project pioneers the concept of networked AUV cluster-based underwater searching system, and utilizes the Software-Defined Networking (SDN) technology to decouple the control plane from the data operation plane of the underwater searching system. On this basis, this project aims to take a deep research on the cooperative path planning mechanism towards the AUV cluster-based underwater searching system from the following three aspects: (1) to ensure the efficiency of information interaction and the control for the AUV cluster, we firstly study the information interaction management scheme based on SDN architecture, when multiple possible scenarios are taken into account; (2) after that, we intend to study the cooperative positioning algorithm and stability optimization model based on adaptive iterative bayesian filtering theory, to improve the localizing accuracy for the AUV cluster; (3) we study the time decay rate restrained GBNN model to improve the real-time ability of path planning towards the AUV cluster. By the aforementioned steps, we realize the efficient, transparent and intelligent cooperative path planning mechanism towards the AUV cluster network-based underwater searching system.