Adaptive Kalman Filtering Based on Model Parameter Ratios
Quanbo Ge, Y. Li, Yuanliang Wang, Xiaoming Hu, H.-T. Li, Changyin Sun
Abstract
This paper studies an adaptive Kalman filter (KF) method based on model parameter ratio (MPR). The model parameter ratio theory is proposed for the first time, and the adaptive estimation problem is transformed into a constrained optimization problem. Compared with the existing Sage-Husa adaptive filtering algorithm, it can be seen that the application of this theory can more accurately estimate the process noise covariance and measurement noise covariance matrix, so that the algorithm has better filtering accuracy and better state estimation performance, At the same time, it is also better in anti divergence and sensitivity to initial conditions.
Topics & Concepts
Kalman filterCovariance matrixDivergence (linguistics)CovarianceControl theory (sociology)Estimation theoryAdaptive filterSensitivity (control systems)Fast Kalman filterNoise (video)Computer scienceInvariant extended Kalman filterExtended Kalman filterAlgorithmMathematicsMathematical optimizationArtificial intelligenceStatisticsEngineeringElectronic engineeringLinguisticsImage (mathematics)PhilosophyControl (management)Target Tracking and Data Fusion in Sensor NetworksInertial Sensor and NavigationMaritime Navigation and Safety