Fiber-Optic Gyroscope Signal Denoising Using an Adaptive Robust Kalman Filter

Abstract

Random noise is an important issue in interferometric fiber-optic gyroscope (IFOG). In this paper, an adaptive robust Kalman filter (KF) and a variant of this are applied to minimize the random noise in IFOG. In the variant of the adaptive robust KF, the measurement noise covariance matrix is adapted using the weighted covariance of the innovation sequence. The suitability of both the algorithms is studied for denoising the IFOG signal under static and maneuvering conditions. In the static case, the Allan variance analysis and the conventional variance are used as the performance indicators to determine the efficiency of the algorithm. In the maneuvering case, root mean-squared error is considered as the performance indicator. The performance of both the algorithms is compared with the conventional KF, innovation-based adaptive estimation adaptive KF, and for minimizing random noise. The experimental results reveal that both the algorithms are competitive for denoising the IFOG signal.