Underwater Data-Driven Positioning Estimation Using Local Spatiotemporal Nonlinear Correlation
Chengming Luo, Luxue Wang, Xudong Yang, Gaifang Xin, Biao Wang · IEEE/CAA Journal of Automatica Sinica · 2023
Dear Editor, A global and local canonical correlation analysis (GLCCA) based on data-driven is presented for underwater positioning. Underwater positioning technology can help the underwater targets move predetermined destinations for specific tasks [1]. Since using different sensor, underwater positioning can be divided into three types: inertial navigation, hydroacoustic positioning and geophysical navigation. Underwater inertial navigation method based on the carried sensors has short-term high accuracy, but it is prone to accumulative errors over time and requires external correction [2]. Within the range of pre-deployed hydroacoustic arrays, hydroacoustic positioning method calculates the signal extracted from nodes, but may have fluctuating errors due to the complex hydroacoustic channels [3]. Geophysical positioning method mainly compares the collected information such as seabed terrain, underwater image, or gravity field with the prior knowledge in the reference database [4]. Different positioning methods have their unique merits as well as inherent drawbacks for different applications. How to combine multiple methods to enhance positioning performance has become the holy grail in the field of underwater positioning.