Tuning kalman filter in linear systems

Kalman filters are used in many different areas that require a solution to discrete-data linear filtering problems. Especially in the field of electric controls, Kalman filters represent a used approach and they are an integral part of many states of the art of electric controls. However, the practical implementation of the Kalman Filter often presents difficulties due to the challenging task of getting a good estimate of the covariance matrix of the process noise and covariance matrix of the measurement noise. A fitting and simultaneous choice of these two matrices based on a feedback loop within the Kalman filter realized by the filter itself can directly lead to an asymptotically stable operating Kalman filter after a reasonable amount of iterations. In this paper an approach to apply a feedback loop enabling dynamic values of the covariance matrix process noise and covariance matrix of the measurement noise is presented. This approach will be applied in simulations using Matlab/Simulink.