基于多AUV的海洋冷涡上升流观测控制策略研究

Aiming for cold eddy-upwelling observation requirements, we design cooperative control and tracking strategy based on the multi AUVs. Firstly, according to the variation characteristics of cold eddy-upwelling profile, we design multi AUVs formation control, so as to adjustment of AUVs coverage range, and make sure that multi AUVs are uniformly distributed and sample in the boundary contour; we control multi AUVs translation and rotation formation, so as to capture and follow drift process of cold eddy-upwelling. Secondly, according variation characteristic of the salinity from the boundary to the center, we control AUV move from the boundary to the center with the lawn-mower path. For the vertical sample, we control AUV to pass through the boundary and the center several times, and we adjust the pitch angle to change the sample resolution. For trajectory design in the horizontal plane, we combine the AUV control and ocean estimation together, so we utilize AUV as the medium to estimate the ocean current, and design the AUV control law with the inversion control algorithm.

本项目针对海洋冷涡上升流动态跟踪观测的需求，设计了基于多水下机器人(AUV)的协同控制和跟踪观测策略。首先，结合冷涡上升流边界处营养盐通量交换特性的研究需求，将冷涡上升流边界视为超椭圆，设计多AUV协同控制队形，使多AUV均布在冷涡上升流的轮廓边界上，进行观测采样；针对轮廓边界的动态漂移、变形现象，控制多AUV队形的移动和旋转，对冷涡上升流的边界进行跟踪观测。其次，针对冷涡上升流从边界处到中心的盐度变化特性的观测需求，控制AUV以梳状锯齿轨迹多次从边界穿越到中心；在剖面锯齿轨迹设计上，根据盐度变化特性，调整AUV俯仰角度，进而动态调整AUV的剖面锯齿轨迹观测采样时的分辨率；观测时AUV以梳状轨迹在水平面运动，针对海流对AUV运动的影响，以AUV作为海流估计的介质，将海流估计和AUV运动控制相结合，设计海流观测器以估算海流大小，并基于反演控制算法设计了AUV跟踪过程中的运动控制算法。

针对海洋冷涡上升流的观测，根据温度标量场及其梯度特性，对其变化特征进行分析、提取，确定了观测区域的大小及边界线。在此基础上，对冷涡上升流的变化特性、跟踪策略进行分析。在固定深度水层的观测上，开展了多AUV协同覆盖观测冷涡上升流区域的算法研究，将多个AUV分布在超椭圆的圆周上，对边界线进行协同覆盖观测，获取观测区域的整体信息；随后，在了解观测区域基本信息的基础上，控制多个AUV均布在冷涡的边界线上，对上升流边界的温度场等值线进行跟踪观测、采样。在剖面观测上，针对AUV剖面俯仰姿态切换的问题，结合已有水下移动观测平台，开展了路径跟踪、动力学建模与控制技术研究，开展水动力计算及系数拟合、动力学建模的研究；结合该平台，基于最优控制和神经网络控制算法，设计AUV俯仰姿态切换的运动控制律，优化了AUV断面观测的俯仰角，抑制海流对运动控制的干扰。