Parameter Estimation of the Hammerstein Output Error Model Using Multi-signal Processing

A parameter estimation method based on multi-signal processing is developed that aims at the Hammerstein output error model in this paper. The multi-signal processing is devised to estimate independently parameters of nonlinear block and linear block for Hammerstein output error model. Firstly, using input-output data of binary signal, the linear block parameters are computed by means of auxiliary model recursive least square method, the unmeasurable variables of the Hammerstein model are effectively handled using auxiliary model technology. In addition, model error probability density function technology is applied to estimate parameters of nonlinear block measurable input-output data of random signal, which not only can control space state distribution of model error, but also make error distribution tend to normal distribution. The results verify that proposed parameter estimation method can effectively estimate the Hammerstein output error model.