Maximum likelihood identification of Wiener-Hammerstein models

The Wiener-Hammerstein (WH) model in discrete form is a cascaded connection of three blocks including a linear dynamics block, a static nonlinearity block, and a (secondary) linear dynamics block. The maximum likelihood (ML) method for SISO- and MISO-WH model parameter identification is applied, and a new ML-computation scheme is proposed, analyzed and tested. The effectiveness and the accuracy of the proposed technique are verified by numerical simulations of mechanical systems including an integrated control valve and heat exchanger system and a superheater-desuperheater in a boiler system.