Hierarchical identification for multivariate Hammerstein systems by using the modified Kalman filter

The parameter estimation problem for multi-input multi-output Hammerstein systems is considered. For the Hammerstein model to be identified, its dynamic time-invariant subsystem is described by a controlled autoregressive model with a communication delay. The modified Kalman filter (MKF) algorithm is derived to estimate the unknown intermediate variables in the system and the MKF-based recursive least squares (LS) algorithm is presented to estimate all the unknown parameters. Furthermore, the hierarchical identification is adopted to decompose the system into two fictitious subsystems: one containing the unknown parameters in the non-linear block and the other containing the unknown parameters in the linear subsystem. Then an MKF-based hierarchical LS algorithm is derived. The convergence analysis shows the performance of the presented algorithms. The numerical simulation results indicate that the proposed algorithms are effective.