哈尔滨工程大学学报2025，Vol.46Issue(4)：627-633,7.DOI:10.11990/jheu.202308026
水下机器人海底地形主动同步定位与建图具身规划算法
Embodiment planning algorithm for AUV bathymetric active SLAM
摘要
Abstract
Addressing the problem of map consistency degradation caused by errors in the inertial navigation sys-tem during autonomous seabed topography scanning missions of autonomous underwater vehicles(AUVs),this paper presents a neuron-excited embodiment planning algorithm for AUV bathymetric active simultaneous localiza-tion and mapping(SLAM).The algorithm facilitates seabed topography matching by revisiting navigable regions and constructing AUV pose constraints,enabling autonomous seabed topography scanning.Within the neural excitation framework,the AUV can recognize and perceive navigable regions and their orientation.This frame-work optimizes the backtracking strategy for the navigable regions in AUV bathymetric active SLAM,autonomous-ly correcting underwater positioning errors.This approach allows for long-time sequence,large-scale underwater accurate navigation and positioning of AUV and the construction of a globally consistent seabed terrain map.Experimental results show that the proposed method effectively solves the problem of active location and embodi-ment planning for AUVs in autonomous seabed terrain scanning,notably improving the scanning range and accu-racy of AUV seabed terrain.关键词
具身路径规划/生物激励神经网络/海底地形测深/主动同步定位与建图/海底地形可导航区/数字地形高程模型/神经元模型/水下机器人Key words
embodiment planning algorithm/biologically inspired neural network/bathymetric survey/active simul-taneous localization and mapping/navigable region/digital elevation model/neuron model/autonomous underwater vehicle分类
信息技术与安全科学引用本文复制引用
张强,游子昂,王建,马腾,李晔,周鑫杰..水下机器人海底地形主动同步定位与建图具身规划算法[J].哈尔滨工程大学学报,2025,46(4):627-633,7.基金项目
国家自然科学基金项目(52431011,52371305). （52431011,52371305）
