Resilient model of multi-UUVs cooperative localization for detection of sensor fault and outliers. (15th December 2022)

• Record Type: Journal Article
• Authors: Du, Zhenqiang; Chai, Hongzhou; Xiang, MinZhi; Zhang, Fan; Wang, Zhaoying
• Is Part Of: Ocean engineering. Volume 266(2022)Part 5
• Publication Date: 2022-12-15
• Subjects: Unmanned underwater vehicle -- Cooperative localization -- Resilient model -- Adaptive factor; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2022.113216
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: Unmanned underwater vehicles (UUVs) cooperative localization technology has become an emerging research topic, which can extend the space-time datum of satellite navigation to seafloor, and provide high-precision location for UUVs to carry out large-area marine environment investigation, exploration and search. Nevertheless, the complex underwater environment makes it impossible to avoid the abnormal condition for sensor fault and outliers. In this contribution, a novel resilient model of multi-UUVs cooperative localization is established. The three-dimensional system model and measurement model are constructed based on the inertial navigation system, and the deviation between multi-sensors are defined and corrected. Furthermore, the resilient factor is proposed based on the normalized prediction residual, and the resilient fusion of UUVs cooperative localization is realized by three steps, i.e., (1) Dynamic resilient fusion of single UUV; (2) Static resilient fusion of single UUV; (3) Static resilient fusion of UUVs cluster. In order to verify the effectiveness of the proposed resilient model, the sound velocity profile in the Nankai trough is selected. The experimental results show that the proposed resilient model of UUVs cooperative localization can accurately detect the abnormal observations and significantly improve the reliability of UUVs cooperative localization. Highlights: A novel resilient model of multi-UUVs cooperative localization is established. The three-dimensional system model and measurement model are constructed. The resilient factor is proposed based on the normalized prediction residual. The proposed model can accurately detect the abnormal observations.