Research On Planning Repair For Autonomous Underwater Vehicle

With the development and utilization of the ocean and the continuous promotion of Autonomous Underwater Vehicle (AUV) application needs, people put forward higher requirements on the level of intelligence and self-handling capacity of AUV. During the execution of AUV mission, because of the uncertainty, such as underwater environment,AUV's own state information and task execution and so on, will generate a large number of uncertain events, once the uncertain events take place, AUV can't guarantee the completion of the task according to the original plan results, it must be according to the existing state and constraint condition to adjust and repair planning. Therefore, in order to improve the autonomy and adaptability of AUV,and to ensure that AUV is safe and effective to complete the mission, AUV should have the ability of planning repair to respond quickly to the environmental change, so that it can adjust to the new environment. This thesis mainly researches on planning repair method for AUV, its main work are as follows:Firstly, this thesis expounds the research status on planning repair technology and its application in AUV,and it sketches the content and significance of research combining the needs of actual implementation mission for AUV.Secondly, for the existing major planning repair methods in the field of intelligent planning is studied; it mainly introduces the summarized ten existing planning repair methods,then the complexity of planning repair problem to carry on the study and analysis.Then, this thesis mainly analyzes the various uncertainties that may be encountered in AUV mission planning, and studies the uncertain event in detail by use some related methods in AI planning. Combined with AUV's own characteristics and the needs of mission planning,this thesis uses two AUV planning repair methods: AUV planning repair based on HTN planning and AUV planning repair based on improved LPA * algorithm; and, we also analyse on the above two planning repair methods and respectively propose the evaluation criteria of the algorithm through research.Finally, for the proposed two AUV planning repair methods in above chapter were verified by experiment, the results show that these methods are effective and feasible; and through the evaluation criteria and comparative analysis of the experimental results proved the superiority of AUV planning repair technology used in this thesis, and also showed that the research contents of this thesis can provide an effective theory and technical support to futher improve the self-adaptability, incident response as well as the safe capability of AUV in the complex underwater environment.