Identification of MISO Wiener and Hammerstein systems

This paper describes an unified new recursive identification method in the prediction error method and model scheme for three MISO Wiener and Hammerstein systems. It is also extension of our earlier work for SISO cases. With the estimation of intermediate variables by using the key term separation principle, a MISO Wiener and Hammerstein system can be approximately transformed into a pseudo-linear MISO dynamic system. Using the adaptive recursive pseudo-linear regressions (RPLR) for a linear MISO dynamic system and smoothing and filtering techniques for estimation of the intermediate variables, satisfied parameter estimates of the MISO Wiener and Hammerstein system can be obtained in the presence of a white or a coloured measurement noise without parameter redundancy. The performance of the developed method is both analysed theoretically and illustrated by means of simulation results.