|
自动化学报 1993
A New Method for the Identification of Hammerstein Model
|
Abstract:
In this paper, a new method for open-loop identification of Hammerstein model is proposed by combining parameter estimation theory with some results in function approximation theory. The new method circumvents the defects of the currently widespread methods for Hammerstein model identification. Using the method, asymptotically unbiased parameter estimates of the model's linear dynamics and the memoryless nonlinearity's approximation polynomial of near the best approximation effect and good convergency can be obtained.