Litcius/Paper detail

A Computationally Efficient Variational Adaptive Kalman Filter for Transfer Alignment

Geng Xu, Yulong Huang, Zhongxing Gao, Yonggang Zhang

2020IEEE Sensors Journal44 citationsDOI

Abstract

To better solve the filtering problem of transfer alignment with an inaccurate measurement noise covariance matrix, a novel computationally efficient version of existing variational adaptive Kalman filter is proposed in this paper, in which an equivalent variational iteration process of the measurement noise covariance matrix is derived. The proposed filter is identical to the existing variational adaptive Kalman filter, but the total computational complexity of algorithm implementation is significantly reduced, which facilitates the application of variational adaptive Kalman filter to transfer alignment. Simulation and experiment results of transfer alignment demonstrate that the computational complexity of the proposed filter is reduced by 55.4% as compared with existing variational adaptive Kalman filter.

Topics & Concepts

Fast Kalman filterInvariant extended Kalman filterKalman filterEnsemble Kalman filterExtended Kalman filterKernel adaptive filterAdaptive filterAlpha beta filterComputer scienceControl theory (sociology)Covariance matrixAlgorithmComputational complexity theoryFilter (signal processing)Noise (video)CovarianceMathematicsFilter designArtificial intelligenceComputer visionMoving horizon estimationControl (management)Image (mathematics)StatisticsInertial Sensor and NavigationTarget Tracking and Data Fusion in Sensor NetworksStructural Health Monitoring Techniques
A Computationally Efficient Variational Adaptive Kalman Filter for Transfer Alignment | Litcius