Influence of Integration Schemes and Maneuvers on the Initial Alignment and Calibration of AUVs: Observability and Degree of Observability Analyses

Author(s)

Frutuoso, Adriano ; Silva, Felipe O. ; de Barros, Ettore A.

• Document type: Journal article
• Source: SENSORS; v. 22, n. 9, p. 21-pg., 2022-05-01.

Abstract

The use of autonomous underwater vehicles (AUV) has increased in a wide range of sectors, including the oil and gas industry, military, and marine research. The AUV capabilities to operate without a direct human operator and untethered to a support vessel are features that have aroused interest in the marine environment. The localization of AUV is significantly affected by the initial alignment and the calibration of the navigation sensors. In this sense, this paper proposes a thorough observability analysis applied to the latter problem. The observability analysis is carried out considering three types of sensor fusion integration and a set of maneuvers, and the results are validated through numerical simulations. As main contribution of this paper, it is shown how the addition of position errors in the observation vector can decouple some gyro and accelerometer biases from the latitude and altitude errors, particularly in the stationary observability analysis. The influence of oscillations in the diving plane and typical AUV maneuvers are analyzed, showing their relative impacts on the degree of observability of the inertial measurement unit (IMU)/Doppler velocity log (DVL) misalignment and DVL scale factor error. Finally, the state's estimation accuracy is also analyzed, showing the limitation of the degree of observability as an assessment tool for the estimability of the states. (AU)

• FAPESP's process: 18/14024-2 - Adaptive sensor fusion applied to AUV navigation
• Grantee: Ettore Apolonio de Barros
• Support Opportunities: Regular Research Grants
• FAPESP's process: 21/13666-3 - Adaptive sensor fusion applied to autonomous underwater vehicle navigation - part 2
• Grantee: Ettore Apolonio de Barros
• Support Opportunities: Regular Research Grants