Auto-calibrated adaptive integrated AHRS/TAM system for orientation estimation of long-range AUVs

Abstract

Accurate, reliable, and real-time orientation estimation is one of the crucial requirements for the safety, performance, and effectiveness of autonomous underwater vehicles (AUVs). Many capabilities in AUVs such as collision-avoidance, trajectory tracking, exploration, and cooperative mission rely heavily on the orientation information including attitude and heading angles. Electromagnetic signal attenuation in the underwater environments as well as time-growing error of the inertial navigation bring about substantial challenges in the orientation estimation of the AUVs. It is more crucial as we use low-cost sensors and technologies for long-term navigation, especially in long-range AUVs. Accordingly, this research is mainly devoted to present an appropriate attitude and heading reference system (AHRS) applied to long-term navigation of the underwater vehicles based on off-the-shelf components. Due to cost constraints, micro-electro mechanical system (MEMS)-grade inertial sensors are used as the inertial measurement unit (IMU) of the proposed navigation system. Considering the above challenges, an auto-calibrated adaptive algorithm is developed for orientation estimation based on decomposed back-stepping (DBS) magnetometer calibration and intelligent fuzzy integration. In the proposed DBS calibration, the accuracy of the traditional magnetic field calibration (MFC) is enhanced through a backward multi-step evaluation-based strategy. In order for better performance, vertical channel and horizontal plane components of the magnetic field vector are decomposed during the evaluation process. In the intelligent fuzzy integration scheme, a Mamdani-based fuzzy inference engine is developed to calculate the maneuvering level of the motion. Consequently, the state-estimation filter is adaptively tuned. The assessment of the proposed low-cost auto-tuned AHRS is conducted through real data obtained in several sea tests.