文献类型：期刊文献

英文题名：Adaptive Zeroing-Gradient Controller for Ship Course Tracking With Near Singularity Considered and Zero Theoretical Tracking Error

作者：Fu, Dongyang[1];Wang, Huan[2];Xiao, Xiuchun[1];Liao, Shan[1];Jin, Long[3]

机构：[1]Guangdong Ocean Univ, Sch Elect & Informat Engn, Zhanjiang 524025, Peoples R China;[2]Guangdong Ocean Univ, Sch Oceanog & Meteorol, Zhanjiang 524025, Peoples R China;[3]Lanzhou Univ, Sch Informat Sci & Engn, Lanzhou 730000, Gansu, Peoples R China

年份：2019

卷号：7

起止页码：38205

外文期刊名：IEEE ACCESS

收录：SCI-EXPANDED(收录号：WOS:000463929000001)、、EI(收录号：20191906896843)、Scopus(收录号：2-s2.0-85065168652)、WOS

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 41340049 and Grant 41430968, in part by the National Marine Important Charity Special Foundation of China under Grant 201305019, in part by the Science and Technology Planning Project of Guangdong under Grant 2013B030200002 and Grant 2016A020222016, in part by the Guangdong University Innovation Outstanding Young Talents Program under Grant 2012WYM-0077, in part by the Project of Enhancing School with Innovation of Guangdong Ocean University under Grant GDOU2014050226, in part by the Key Lab of Digital Signal and Image Processing of Guangdong Province under Grant 2016GDDSIPL-02, in part by the Doctoral Initiating Project of Guangdong Ocean University under Grant E13428, in part by the Fundamental Research Funds for the Central Universities under Grant lzujbky-2017-193, and in part by the Project of Enhancing School with Innovation of Guangdong Ocean University under Grant Q15090.

语种：英文

外文关键词：Zeroing dynamics; gradient dynamics; adaptive method; activation function; ship course tracking

外文摘要：The zeroing-gradient (ZG) method, a combination of two neural dynamics methods, has been leveraged to construct controller for solving ship course tracking problem with near singularity considered. In this paper, the existing ZG controller is presented and revisited with its limitation identified. Specifically, this paper points out that there exist lagging errors for the ZG controller with the theoretical analyses provided. Then, to remedy the weakness, a new controller is proposed with an adaptive coefficient with zero theoretical tracking errors. Moreover, different activation functions are exploited to accelerate the convergence of the proposed controller. Finally, simulations are conducted to verify the superiority of the proposed controller.