Recursive parameter estimation under lack of excitation
Transient and stationary properties of a recently suggested by B. Stenlund and F. Gustafsson (covariance) anti-windup scheme in the Kalman Filter (KF) are studied. It appears that the Riccati equation of the KF can be for this special case reduced to a Sylvester equation which allows for linear analysis. Convergence of parameter estimates under full as well as insufficient excitation is studied together with effective computational implementations in high order adaptive filters. Applications to active vibration control, echo cancellation and fault detection are envisioned.
The project is financed by Vetenskapsrådet.
- A. Medvedev and M. Evestedt, “Elementwise decoupling and convergence of the Riccati equation in the SG-algorithm,” to be presented at 17th IFAC World Congress, Seoul, Korea, July 2008.
- M. Evestedt and A. Medvedev, “Recursive parameter estimation by means of the SG-algorithm,” to be presented at 17th IFAC World Congress, Seoul, Korea, July 2008.
- M. Evestedt, A. Medvedev and T. Wigren, “Windup properties of recursive parameter estimation algorithms in acoustic echo cancellation”, accepted for publication in Control Engineering Practice, 2008
- M. Evestedt and A. Medvedev, Stationary behavior of an anti-windup scheme for recursive parameter estimation under lack of excitation, Automatica, v. 42 , no. 1, pp. 151--157, January, 2006.
- A. Medvedev, "Stability of a Riccati Equation Arising in Recursive Parameter Estimation Under Lack of Excitation", IEEE Transactions on automatic control, v. 49, pp. 2275--2280, December, 2004