A target tracking method with colored measurement noise and variational Bayesian adaptive Kalman filter

Abstract

The invention belongs to the technical field of carrier navigation of ships, aircraft, vehicles and the like, in particular to a target tracking method with colored measurement noise and variational Bayesian adaptive Kalman filter. Including: 1. Establishing the state equation and measurement equation of target tracking. 2. converting that color measurement noisento white measurement noise by themeasurement difference method. 3. SelectingThe prior distribution of the one-step prediction covariance matrix and the measurement covariance matrix of the state spread vectors the inverse Wishart distribution. 4. Performing Variational approximation of joint posterior probability density function. 5. Estimating The extended state vector and its one-step prediction covariance matrix and measurement covariance matrix by variational Bayesian method. The method of the invention completes the state estimation task in the target tracking process under the condition of imprecise noise covariance matrix and colored measurement noise, and the tracking accuracy is higher than the existing target tracking method based on other filters.

Description

technical field [0001] The invention belongs to the technical field of navigation of ships, airplanes, vehicles and the like, and in particular relates to a target tracking method with colored measurement noise and variational Bayesian adaptive Kalman filtering. Background technique [0002] The Kalman filter has been widely used in many engineering applications, where the Kalman filter assumes that the noise covariance matrix is ​​known exactly. However, in many applications with imprecise noise covariance matrices, the performance of the Kalman filter may degrade. Adaptive Kalman filter based on variational Bayesian is an excellent solution to this problem. By choosing an appropriate conjugate prior distribution, the existing variational Bayesian adaptive Kalman filter jointly estimates the state vector, the imprecise and slowly changing system noise covariance matrix and the measurement noise covariance matrix. Since existing variational Bayesian-based adaptive Kalman f...

AI Technical Summary

Problems solved by technology

By choosing an appropriate conjugate prior distribution, the existing variational Bayesian adaptive Kalman filter jointly estimates the state vector, the imprecise and slowly changing system noise covariance matrix and the measurement noise covariance matrix

Since existing variational Bayesian-based adaptive Kalman filters are specifically designed for white measurement noise, the performance of existing methods may degrade for colored measurement noise

[0003] Although the problem with colored measurement noise and imprecise system noise covariance matrices and measurement noise covariance matrices can be transformed into a system noise covariance matrix with one-step delay state and imprecise quantities using existing measurement-difference methods The problem of measuring the noise covariance matrix, but it will bring two problems

First of all, the measurement of the state space model constructed after measuring the difference at the current moment depends not only on the current state, but also on the state at the previous moment, and the existing variational Bayesian adaptive Kalman filter cannot be used for dealing with the problem of linear state-space models with one-step delayed states; secondly, in the process of updating the one-step forecast error covariance matrix of the extended state vector, not only the one-step forecast error covariance matrix needs to be estimated, but also the estimate from the previous moment The error covariance matrix, so the existing variational Bayesian adaptive Kalman filter cannot estimate the one-step forecast error covariance matrix of the extended state vector

Images