Integrated navigation algorithm for deep-sea AUV based on M-estimation and heavy-tailed noise
Haosu Zhang,Cheng Ma,Chaoqi Chen,Shiyin Zhou,Zhiying Wei,Ruohan Zheng,Tao Wang,Jianming Miao,Jinxin Shao,Xi Gong,Xingyu Sun
DOI: https://doi.org/10.1016/j.apor.2023.103781
IF: 3.761
2023-11-05
Applied Ocean Research
Abstract:This paper presents an integrated navigation algorithm for deep-sea Autonomous Underwater Vehicles (AUVs) based on M-estimation and heavy-tailed noise to solve the problems of inaccurate noise models in traditional navigation algorithms for deep-sea AUVs, significant delays in Ultra-Short Baseline (USBL) acoustic positioning, and the inability to estimate deep-sea ocean currents. The algorithm encompasses four techniques: 1A Strapdown Inertial Navigation Systems (SINS)/USBL integrated navigation algorithm based on M-estimation and heavy-tailed noise models. 2Estimation and compensation of USBL acoustic delay positioning errors. 3Estimation and compensation of steady or slowly varying ocean currents. 4Considering the distance between the submerged vehicle and the center of the earth and the variation of gravity acceleration with depth. The first and second techniques are the innovations of this article, which can effectively improve the USBL calibration accuracy with only a slight increase in computational complexity. The significant positioning error caused by long-term voyages of large-depth AUVs in the middle layer of the ocean can be effectively suppressed using the third and fourth technologies. The results of the semi-physical simulation and lake experiment demonstrate the ability of the proposed algorithm to significantly improve navigation accuracy and robustness in complex underwater environments compared to conventional methods.
engineering, ocean,oceanography