基于动态反馈的AUV水平面路径跟踪控制

针对离散自治水下机器人水平面的路径跟踪控制问题，利用水下机器人的位置和姿态角量测信息，提出神经网络自适应输出反馈控制器.所设计的控制器包括3部分，镇定水下机器人动态系统线性部分的动态反馈控制器；神经网络控制器，用来补偿水下机器人受到环境干扰引起的水动力系数变化的不确定非线性结构；补偿环境扰动和神经网络带来的重构误差的鲁棒控制器.基于离散非线性系统理论，对水下机器人水平面跟踪误差系统进行分析，得到系统的跟踪误差最终一致有界.所提出的控制方法具有对模型精确度要求低的优点，利用INFANTE水下机器人模型进行仿真分析验证了提出的控制算法的有效性.
Abstract: Aiming at the path following control in horizontal plane of discrete-time autonomous underwater vehicle (AUV), a neural network adaptive output feedback controller was proposed using the information of AUV position and attitude angle. The controller consisted of three parts:dynamic feedback controller which was used to stabilize the linear part of AUV dynamic system; neural network controller which was used to compensate the AUV nonlinear uncertainty induced by hydrodynamic coefficients; and robust controller which was used to compensate the reconstructive error of neural network and environmental disturbances. Based on the theory of discrete-time nonlinear systems, the AUV tracking error system in horizontal plane was analyzed and proved to be uniformly ultimately bounded. The proposed control algorithm required low accuracy of the system model. The model of INFANTE AUV was taken to verify the effectiveness of the proposed control algorithm by simulation.