Variational Bayesian self-adaptive filtering method

Abstract

The invention relates to a variational Bayesian self-adaptive filtering method. For the target status estimation problem in a case of unknown process noise time-varying, inverse Wishart distribution is used for a priori modeling on process noise covariance through introducing a new hidden variable, and target status, the process noise covariance and the hidden variable are jointly optimized undera variational Bayesian framework through an iterative mechanism to improve tracking accuracy of the target status. Compared with the existing filtering algorithms, the method of the invention directlyselects the conjugate a priori distribution of the process noise covariance, carries out iterative optimization of a target status estimation result under the variational Bayesian framework, obtainsa new self-adaptive filtering algorithm, and has certain theoretical and practical engineering significance for the field of filters.

Description

technical field [0001] The invention belongs to a filter subsystem and relates to a variational Bayesian adaptive filtering method. Background technique [0002] The classical Kalman filter algorithm needs to know the system process noise and measurement noise when performing state estimation. In actual systems, such as target tracking systems, there are uncertainties in the dynamic evolution of the system, and the time-varying process noise is unknown. The filtering algorithm will cause the estimation error to become larger and even diverge. The adaptive filtering algorithm based on variational Bayesian can estimate the system model parameters while estimating the state, but its conjugate prior model is needed to obtain the analytical solution when modeling the parameters. For the unknown process noise covariance, since the process noise covariance appears in the calculation of the state prediction covariance in the Kalman filter formula, its conjugate prior cannot be simp...

AI Technical Summary

Problems solved by technology

[0002] The classical Kalman filter algorithm needs to know the system process noise and measurement noise when performing state estimation. In actual systems, such as target tracking systems, there are uncertainties in the dynamic evolution of the system, and the time-varying process noise is unknown. The filtering algorithm will cause the estimation error to become larger and even diverge

The adaptive filtering algorithm based on variational Bayesian can estimate the system model parameters while estimating the state, but its conjugate prior model is needed to obtain the analytical solution when modeling the parameters.

For the unknown process noise covariance, since the process noise covariance appears in the calculation of the state prediction covariance in the Kalman filter formula, its conjugate prior cannot be simply obtained. Other algorithms based on variational Bayesian Either transform it into a smoothing problem and process it offline, or transform the problem of unknown covariance of process noise into a problem of unknown predictive covariance. In the process of filtering implementation, an initial given process noise covariance is required at each moment

Images