M. Ruderman and T. Bertram (Germany)
Preisach model, Density-function estimation, Hysteresis, Non-Parametric Identiļ¬cation, Nonlinear operator
Preisach hysteresis model provides a powerful means to describe arbitrary hysteresis effects with a rate-independent behavior. When using the Preisach model a density func tion of spatial distributed elementary hysteresis operators has to be found from experimental data which are noisy and often limited. In this paper, we propose a novel estimation technique to identify the discrete Preisach density function from a fairly small data set. The method involves a nonlinear least-squares algorithm in order to minimize a vector-valued function which is derived with respect to the model properties. We show a good agrement between the model prediction and the input-output behavior of the hysteresis system for both simulated and experimental data.
Important Links:
Go Back
