Multi-AUVs Cooperative Complete Coverage Path Planning Based on GBNN Algorithm

Daqi Zhu,Chen Tian,Xiaofei Jiang,Chaomin Luo
DOI: https://doi.org/10.1109/ccdc.2017.7978395
2017-01-01
Abstract:For the complete coverage path planning of multi-AUVs (autonomous underwater vehicles), a method of cooperative complete coverage path planning for multi-AUVs based on GBNN algorithm is proposed in this paper, where all AUVs share the common working environment information, and each AUV treats other AUVs as moving obstacles. Firstly, a grid map is constructed by discretizing the underwater workspace. Secondly, the corresponding dynamic neural networks are built on the grid map, and there are N neural networks for N AUVs, respectively. Finally, each AUV plans its reasonable coverage path according to its own neural activity through GBNN strategy, until work together to cover the entire workspace. The simulation results show that multi-AUVs can not only cooperatively accomplish the common coverage tasks without any mutual collision and overlapped coverage, but also automatically avoid obstacles and escape from deadlocks, and effectively improve work efficiency of each single AUV, reduce complete coverage time and save AUVs' power.
What problem does this paper attempt to address?