In the complex ocean environment, the autonomous underwater vehicle (AUV) is subject to the multiple uncertainties including deterministic uncertainties and stochastic uncertainties, which greatly degrade the control performance. To solve this problem, a novel robust control strategy using combined uncertainty and disturbance estimator (UDE) and unscented Kalman filter (UKF) is proposed for the path-following control (PFC) of the underactuated AUV under multiple uncertainties in this paper. First, the UDE technique is adopted to handle the deterministic uncertainties, including deterministic components of environmental disturbances, parameter uncertainties, and unmodeled dynamics, etc. Second, the uncertainty of unactuated lateral channel, which cannot be dealt with by the UDE, is treated as an unknown parameter. Also, to tackle the stochastic uncertainties, which is often ignored in previous studies, including stochastic components of environmental disturbances, measurement noise, and the estimation errors of UDE which are treated as the process noise, the augmented UKF technique is adopted to jointly estimate the system states and the lateral channel unknown parameter. Finally, extensive numerical simulations and comparative analyses are presented to verify the efficiency and robustness of the proposed control strategy.

在复杂的海洋环境中，自主水下航行器（AUV）受到确定性不确定性和随机性不确定性等多重不确定性的影响，极大地降低了控制性能。为了解决这个问题，本文提出了一种新颖的鲁棒控制策略，该策略使用组合不确定性和扰动估计器（UDE）和无迹卡尔曼滤波器（UKF），用于多重不确定性下欠驱动AUV的路径跟踪控制（PFC）。首先，采用UDE技术来处理确定性不确定性，包括环境扰动的确定性成分、参数不确定性和未建模动力学等。其次，UDE无法处理的非驱动横向通道的不确定性被视为未知参数。还，为了解决以往研究中经常被忽视的随机不确定性，包括环境扰动的随机成分、测量噪声以及被视为过程噪声的UDE估计误差，采用增强UKF技术来联合估计系统状态和横向通道未知参数。最后，进行了广泛的数值模拟和比较分析，以验证所提出的控制策略的效率和鲁棒性。