Research On The Environment Modeling Method Of The AUV Path Planning System

The environment modeling is crucial to the path planning of autonomous underwatervehicle. It is inevitable for underwater vehicle to be affected by the complex oceanenvironment. Traditional environment models usually just take known information about theocean terrain into account and cannot entirely reflect the actual ocean environment condition,which are not completely for the available information of the path planning. Therefore, inorder to better assist underwater vehicle in path planning, it is urgently necessary to describethe ocean environment completely and accurately.The paper focuses on the application requirements of the ocean environment informationfor the AUV path planning and provides an ocean environment modeling method whichsynthesizes the ocean terrain, the ocean hydrologic element and the sensing information ofsonar sensor. This method is able to achieve a complete and accurate description of the oceanenvironment information.In terms of the modeling of the ocean terrain, considering the low efficiency of theterrain modeling method based on the basic Delaunay triangulated irregular network, thepaper offers a compositive Delaunay triangulated network algorithm based on the divide andconquer algorithm and the incremental insertion algorithm, and then applies this new methodto modeling the ocean terrain.In terms of the modeling of the ocean hydrologic element, the paper mainly conducts theresearch in three aspects. First, we analyze the effects made by ocean current, tide current,thermocline，sound channel，transparence and other ocean hydrologic element on theperformance of autonomous underwater vehicle, and proposes a regulation to identify the safenavigation region of AUV. Second, we present a method for ranking the ocean hydrologicelements according to the need of path planning for underwater vehicle. Finally, we refer tothe attributes of all kinds of ocean hydrologic element and establish the relevant data storagemodel in NetCDF format. In addition, we also establish the data model applied to the pathplanning of AUV. These methods can enhance the efficiency of data retrieval, thus serving theAUV path planning system better.In terms of the model of the sonar sensor sensing information, the paper carries outresearch on the sensing environment information gained by sonar sensor and the uncertaintyof the information. We describe the probability distribution of the environment informationbased on the probability theory, and establish the information model, which improving the accuracy of the description for sonar sensor sensing information. In the end, we acquire theenvironment model of the corresponding area based on the sonar sensor sensing information.Lastly, In order to provide more complete and more authentic environment informationsupport to the path planning of underwater vehicle, based on the ocean terrain model and theocean hydrological element model, we attain the design of the safe surface which is propitiousto the path planning of autonomous underwater vehicle by the regulation of identifying thesafe navigation region of AUV and the precedence relationship of all kinds of oceanhydrological element.