小型吊舱式无人艇航向控制  

The Course Control of Small Pod-Type Unmanned Surface Vehicle Using RBF Neural Network Adaptive Sliding Mode Control

作　　者：郑艳芳 俞万能[1,2,3] 廖卫强 蒋仁炎[1] ZHENG Yanfang;YU Wanneng;LIAO Weiqiang;JIANG Renyan(School of Marine Engineering,Jimei University,Xiamen 361021,China;Fujian Province Key Laboratory of Naval Architecture and Ocean Engineering,Xiamen 361021,China;National and Local Joint Engineering Research Center for Ship Aided Navigation Technology,Xiamen 361021,China)

机构地区：[1]集美大学轮机工程学院,福建厦门361021 [2]福建省船舶与海洋工程重点实验室,福建厦门361021 [3]船舶辅助导航技术国家地方联合工程研究中心,福建厦门361021

出　　处：《集美大学学报（自然科学版）》2022年第1期55-62,共8页Journal of Jimei University：Natural Science

基　　金：国家自然科学基金项目(51679106,52171308)。

摘　　要：针对小型吊舱式无人艇航向控制系统精度问题,考虑模型中的不确定性和风、浪干扰等未知项,设计一种基于RBF神经网络和迭代滑模算法的自适应控制器。在建立吊舱式无人艇运动数学模型基础上,采用迭代滑模算法提高收敛时间,并通过RBF神经网络权值逼近模型参数不确定项和未知扰动,最终将该算法与迭代滑模算法进行仿真比较。结果表明,所提出的自适应控制算法可减弱迭代滑模抖振现象,提高收敛速度和航向控制精度,满足无人艇对航向偏差控制的要求。Aiming at the accuracy problem on the course control of small pod-type USV,while considering the uncertainties of the model by the unknown terms such as wind and wave interferences,an adaptive controller based RBF neural network and iterative sliding mode algorithm is developed.On the basis of establishing the mathematical model for the motion of the pod-type USV,an iterative sliding mode algorithm was used to shorten the convergence time,and the RBF neural network weights were employed to approximate the uncertain items of model parameters and unknown disturbances.Finally,comparison of course control stability was carried out between iterative sliding mode algorithm and this algorithm.The results show that the proposed adaptive control algorithm can weak the iterative sliding mode chattering phenomenon,improve the convergence speed and the course control accuracy,and meet the requirements of the USV for course deviation control.

关 键 词：小型吊舱无人艇 航向控制 RBF神经网络 迭代滑模算法 

分 类 号：U664.82[交通运输工程—船舶及航道工程]