Qiu Jianbin, Ma Min, Wang Tong, Gao Huijun
IEEE Trans Neural Netw Learn Syst. 2021 Dec;32(12):5266-5273. doi: 10.1109/TNNLS.2021.3056585. Epub 2021 Nov 30.
This article investigates the adaptive learning control problem for a class of nonlinear autonomous underwater vehicles (AUVs) with unknown uncertainties. The unknown nonlinear functions in the AUVs are approximated by radial basis function neural networks (RBFNNs), in which the weight updating laws are designed via gradient descent algorithm. The proposed gradient descent-based control scheme guarantees the semiglobal uniform ultimate boundedness (SUUB) of the system and the fast convergence of the weight updating laws. In order to reduce the computational burden during the backstepping control design process, the command-filter-based design technique is incorporated into the adaptive learning control strategy. Finally, simulation studies are given to demonstrate the effectiveness of the proposed method.
本文研究了一类具有未知不确定性的非线性自主水下航行器（AUV）的自适应学习控制问题。AUV中的未知非线性函数由径向基函数神经网络（RBFNN）逼近，其中权重更新律通过梯度下降算法设计。所提出的基于梯度下降的控制方案保证了系统的半全局一致最终有界性（SUUB）以及权重更新律的快速收敛。为了减轻反步控制设计过程中的计算负担，将基于指令滤波器的设计技术纳入自适应学习控制策略。最后，通过仿真研究验证了所提方法的有效性。
