An internal model approach to the identification of time-varying systems.
G. Davidov, A. Shavit, N. Shirnkhi, Yoram Koren · 2005
The identification of linear SISO stable discrete systems with dynamical-varying parameters is treated. An algorithm, based on the Internal Model principle and on the recursive Least Squares (IMLS) for identification of such systems is proposed. It is assumed that the system parameters satisfy known linear difference equations with constant coefficients. A persistent excitation condition of the measurement vector, automatically guarantees exponential stability and therefore there is no need to use any resetting procedure as was done in some previous works. Simulation results reveal good tracking even of parameters that vary drastically in an abrupt manner.