A Vision Aided Initial Alignment Method of Strapdown Inertial Navigation Systems in Polar Regions

MEMS strapdown inertial attitude measurement system using rotational modulation technology

Attitude determination involves the integration of methodologies and systems for estimating the time varying attitude of moving objects. Strapdown Inertial Attitude Measurement System (SIAMS) is among the most widely used navigation systems. The development of cost effective Micro Electro Mechanic System (MEMS) based inertial sensors has made attitude measurement system more affordable. However, MEMS sensors suffer from various errors that have to be calibrated and compensated to get acceptable attitude results. Given the auto-compensation of inertial sensor bias in rotation error modulation, the objective of this paper is to develop a MEMS-based rotary SIAMS, in which the significant sensor bias is automatically compensated by rotating the IMU, to offer comparable performance with respect to a tactical-grade Inertial Measurement Unit (IMU). With the analysis of the relationship between the MEMS error and misalignment, a MEMS calibration model is derived, and a combined calibration method of multi position rotation is applied to estimate the deterministic sensor errors such as bias, scale factor, and misalignment. Simulation and experiment results indicate that the proposed method can further modulate and compensate the MEMS errors, thereby improving the MEMS attitude accuracy.

A Polar Robust Kalman Filter Algorithm for DVL-Aided SINSs Based on the Ellipsoidal Earth Model

Autonomous underwater vehicles (AUVs) play an increasingly essential role in the field of polar ocean exploration, and the Doppler velocity log (DVL)-aided strapdown inertial navigation system (SINS) is widely used for it. Due to the rapid convergence of the meridians, traditional inertial navigation mechanisms fail in the polar region. To tackle this problem, a transverse inertial navigation mechanism based on the earth ellipsoidal model is designed in this paper. Influenced by the harsh environment of the polar regions, unknown and time-varying outlier noise appears in the output of DVL, which makes the performance of the standard Kalman filter degrade. To address this issue, a robust Kalman filter algorithm based on Mahalanobis distance is used to adaptively estimate measurement noise covariance; thus, the Kalman filter gain can be modified to weight the measurement. A trial ship experiment and semi-physical simulation experiment were carried out to verify the effectiveness of the proposed algorithm. The results demonstrate that the proposed algorithm can effectively resist the influence of DVL outliers and improve positioning accuracy.