摘 要： AUV作为人类开发探索海洋的重要工具拥有不可替代的作用，但是单个AUV有时很难完成复杂的任务，因此具有良好协作性和互补性的多AUV系统逐渐受到广泛关注。多AUV在水下可以形成联合围捕的规模，通过相互通信相互协调，应用多种围捕策略对目标进行控制。本文以多AUV在水下进行协同围捕单个目标的案例为基础进行了深入的研究和讨论，对多AUV协同围捕的算法策略进行了理论研究与仿真实验，通过生物启发神经网络方法使得AUV在航行路线中对神经元的输出值进行最优选择，从而更快更直接逼近目标，同时结合协商分配围捕点的办法，使得AUV能够在各自最短的路径内占据既定围捕点，并列举了在无障碍物、静态和动态障碍物环境下的仿真，验证了生物启发神经网络算法的优越性和实用性。另外为了克服海流对AUV航行的影响，提出方向决策的算法，使得AUV能够在预计的方向移动。93751

毕业论文关键词：AUV；协同围捕；矢量合成；生物启发神经网络

Abstract: AUV as an important tool of human development to explore the ocean has an irreplaceable role, but a single AUV is sometimes difficult to complete complex tasks, multi AUV system, so it has good cooperative and complementary gradually attracted widespread attention。 Multi AUV can form a joint round up under water, through mutual communication and coordination, through a variety of strategies to control the target。 The AUV in the water conducted in-depth study and discussion is carried out on the basis of cooperative hunting single target case, the algorithm of multi AUV cooperative hunting strategy has carried on the theoretical research and simulation experiment, the biologically inspired neural network method makes AUV optimal selection of element values in the output of the neural navigation route, thus faster a more direct approach target, combined with the distribution of the ways round negotiation, so that the AUV can occupy the fixed points in their respective rounding up the shortest path, and listed in the simulation without barriers, static and dynamic obstacle environment, verifies the practicability and superiority of the biologically inspired neural network algorithm。 In addition, in order to overcome the influence of ocean current on AUV navigation, an algorithm of direction decision is proposed, which can make AUV move in the expected direction。

Key words: AUV; Cooperative rounding；Vector composition；Biologically inspired neural network

目  录

1 绪论 4

1。1研究背景 4

1。2 多AUV围捕控制概述 5

1。3多AUV围捕控制研究方法 5

1。4本文研究内容 6

2  基于生物启发神经网络围捕算法研究 6

2。1  生物启发神经网络的基本原理 6

2。2  生物启发神经网络模型 7

2。3  模型稳定性分析及相关参数的敏感性讨论 9

2。3。1  二维神经网络模型稳定性分析 9

2。3。2  相关参数的敏感性讨论 9

3  二维水下环境中多AUV围捕研究 13

3。1  二维水下环境中围捕描述 13

3。2  目标AUV逃逸运动研究 15

3。3  围捕策略研究 16

3。4  二维水下环境中围捕任务仿真