CAJ | 학술논문

自主水下机器人(Autonomous Underwater Vehicle,AUV)是海洋开发与探索的有效工具。为提高 AUV在复杂海洋环境、任务多变以及通信受限等不确定条件下的自适应性和任务执行的可靠性,首先,研究并设计了基于分层思想的 AUV任务规划与重规划体系结构；其次,针对不同的层次分别提出了基于与或分解树的使命规划、基于有限状态机的任务规划；然后,阐述了分层重规划的意义并设计了分层重规划监督决策的具体算法；最后,仿真实验表明了所设计的分层体系结构及分层任务规划与重规划监督决策算法,能显著提高 AUV不确定条件下的自适应性和自主完成任务可靠性。
자주수하궤기인(Autonomous Underwater Vehicle,AUV)시해양개발여탐색적유효공구。위제고 AUV재복잡해양배경、임무다변이급통신수한등불학정조건하적자괄응성화임무집행적가고성,수선,연구병설계료기우분층사상적 AUV임무규화여중규화체계결구；기차,침대불동적층차분별제출료기우여혹분해수적사명규화、기우유한상태궤적임무규화；연후,천술료분층중규화적의의병설계료분층중규화감독결책적구체산법；최후,방진실험표명료소설계적분층체계결구급분층임무규화여중규화감독결책산법,능현저제고 AUV불학정조건하적자괄응성화자주완성임무가고성。
Autonomous underwater vehicle(AUV)is an effective equipment for ocean development and exploration.It is important to improve the adaptability and reliability of AUV mission execution in consideration of various uncertainties such as complex ocean environment,changeable mission and limited communication.In this paper,we study and design a hierarchical mission planning and re-planning architecture.It is based on hierarchical task network which can describe experiential knowledge and solve problems efficiently.Secondly,we propose the mission planning method based on and-or decomposition tree and task planning scheme based on the finite state machine respectively. These methods achieve the hierarchical re-planning model and handle different uncertainty on different level.Then, the significance of hierarchical mission re-planning is expounded and the supervision and decision-making algorithm for hierarchical mission re-planning is presented.Finally,simulation experiments indicate that the designed hierarchical architecture,mission planning and re-planning supervision and decision-making methods can significantly improve the adaptability and reliability of AUV mission execution.