Research on autonomous planning for AUV in unstructured environment

Hongjian Wang,Dehui Zhao,Xinqian Bian,Xiaocheng Shi
DOI: https://doi.org/10.1007/11816157_71
2006-01-01
Abstract:Autonomous planning is an important ability for autonomous underwater vehicle (AUV), which is a crucial factor for ensuring and guiding AUV to long-term navigate and accomplish its mission in large range and unstructured oceanic area. In general, the technology of autonomous planning could be sorted into global path planning and local path planning. In this paper, the former is solved by adopting adaptive genetic algorithm (AGA), which aimed at searching an optimized path according to some optimization criteria in a known and certain environment. This algorithm also can be adapted and applied to dynamic plan a collision-free path based on sonar and real-time navigate AUV in unknown and uncertain environment. By analyzing the motion security of AUV, two kinds of regions which are defined as forbidden zone and potential collision zone, and some safety motion criterions are also proposed for AUV. All of above researches are simulated in a semi-physical simulation platform, and the results show that the autonomous planning algorithms are valid, efficient and reliable. The adaptive abilities and autonomy of AUV can make it possible to long-term navigate in ocean and succeed in its mission.
What problem does this paper attempt to address?