文献类型：期刊文献

英文题名：Multi-joint adaptive control enhanced reinforcement learning for unmanned ship

作者：Li, Jiawen[1,2,3,4];Jiang, Xin[1];Zhang, Hao[1];Wu, Langtao[1];Cao, Liang[1,2,3,4];Li, Ronghui[1,3,4]

机构：[1]Guangdong Ocean Univ, Naval Architecture & Shipping Coll, Zhanjiang 524005, Peoples R China;[2]Hainan Vocat Univ Sci & Technol, Key Lab Philosophy & Social Sci Hainan Prov Hainan, Haikou 570100, Peoples R China;[3]Tech Res Ctr Ship Intelligence & Safety Engn Guang, Zhanjiang 524005, Guangdong, Peoples R China;[4]Guangdong Prov Key Lab Intelligent Equipment South, Zhanjiang 524005, Guangdong, Peoples R China

年份：2025

卷号：318

外文期刊名：OCEAN ENGINEERING

收录：SCI-EXPANDED(收录号：WOS:001394880400001)、、EI(收录号：20245217591578)、Scopus(收录号：2-s2.0-85212918722)、WOS

基金：This research was funded by the National Natural Science Foundation of China (Grant No. 52171346) , the Ocean Young Talent Innovation Programme of Zhanjiang City (Grant No. 2022E05002) , the Young Innovative Talents Grants Programme of Guangdong Province (Grant No. 2022KQNCX024) ,the special projects of key fields of Universities in Guangdong Province (Grant No. 2023ZDZX3003) , The China Institute of Navigation Young Elite Scientist Sponsorship Program by CIN (Grant No. YESSCIN2023008) , the College Student Innovation Team of Guangdong Ocean University (Grant 680No. CXTD2024018) , the Natural Science Foundation of Guangdong Province (Grant No. 2021A1515012618) , the program for scientific research start-up funds of Guangdong Ocean University, the China Transportation Education Research Association (Grant No. JTYB20-28) , the Guangdong Provincial Education Teaching Reform Research Project (Grant No. 010202132201) , Zhanjiang Social Science Association (Grant No. ZJ20YB0) .

语种：英文

外文关键词：Autonomous navigation; Unmanned ship; Reinforcement learning; Rudder control; Route optimization; Safety performance

外文摘要：The pursuit of stable and reliable autonomous navigation is a prevailing research focus within the domain of Unmanned Ship technology. This study introduces the Multi-Joint Adaptive Control Enhanced Reinforcement Learning System (MAES) that enhances the autonomous stability of unmanned vessel navigation in maritime settings. The MAES architecture integrates a digital platform module, an auto-learning driving module, and a rudder position optimization module, thereby constructing a comprehensive autonomous ship navigation model that accounts for rudder angle control and course discrepancies. Empirical navigation experiments conducted across eight virtual waterways have demonstrated the MAES significantly reduces the number of steering maneuvers required to reach the destination when juxtaposed with conventional autonomous navigation models. Furthermore, the MAES exhibits commendable route-search efficiency, achieving an average reduction of 89.29 units in route length as compared to other reinforcement learning models. In terms of safety performance, the MAES model has demonstrated a considerable reduction in potential risks, averaging a 22.2% decrease when contrasted with traditional autonomous navigation model.