Data filtering based recursive least squares algorithm for Hammerstein systems using the key-term separation principle

This paper concerns parameter identification of Hammerstein output error moving average systems with a two-segment piecewise nonlinearity. By combining the key-term separation principle and the data filtering technique, we transfer the Hammerstein model into two regression identification models, and present a data filtering based recursive least squares method to estimate the parameters of these two identification models. The proposed algorithm achieves a higher computational efficiency than the standard approach by using covariance matrices of smaller dimensions from the two identification models instead of one identification model in the standard approach. (C) 2012 Elsevier Inc. All rights reserved.