Hammerstein system identification using robust estimator based on quantized observation

Hammerstein system is the most popular block-oriented model, which can represent a large number of nonlinear model features. With the rapid development of science and technology, quantization communication has become a hot topic in data transmission field based on quantized sensor. Because Hammerstein system can describe an actuator in series with linear system, the quantized Hammerstein system are receiving more and more attention. In this context, this work discusses the identification of the Hammerstein system in the presence of quantitative observation. To obtain the effective system data, a constant filter gain is used to filter the system input and output data. Then, the multi innovation theory is introduced to expand the prediction error data into matrix form. Based on the above matrices, the augmented parameter error data can be gained. Right after, the robust estimator is given by using the augmented parameter error data. Finally, we apply the quantized example to examine the performance of the proposed estimator.