This paper considers the estimation of the process state and noise parameters when the statistics of the process and measurement noise are unknown or time varying in the integration system. An adaptive Kalman Filter (AKF) with a recursive noise estimator that is based on maximum a posteriori estimation and one-step smoothing filtering is proposed, and the AKF can provide accurate noise statistical parameters for the Kalman filter in real-time. An exponentially weighted fading memory method is introduced to increase the weights of the recent innovations when the noise statistics are time varying. Also, the innovation covariances within a moving window are averaged to correct the noise statistics estimator. Experiments on the integrated Strapdown Inertial Navigation System (SINS)/ Doppler Velocity Log (DVL) system show that the proposed AKF improves the estimation accuracy effectively and the AKF is robust in the presence of vigorous-manoeuvres and rough sea conditions.
Adaptive Kalman Filtering with Recursive Noise Estimator for Integrated SINS/DVL Systems
The journal of navigation ; 68 , 1
2015
Article (Journal)
English
Adaptive Two-stage Kalman Filter for SINS/Odometer Integrated Navigation Systems
Online Contents | 2017
|Adaptive Two-stage Kalman Filter for SINS/Odometer Integrated Navigation Systems – ERRATUM
Online Contents | 2017
|Adaptive Two-stage Kalman Filter for SINS/Odometer Integrated Navigation Systems – ERRATUM
Online Contents | 2016
|Adaptive Two-stage Kalman Filter for SINS/Odometer Integrated Navigation Systems – ERRATUM
Online Contents | 2016
|