Identification of sandwich systems with hysteresis based on two-stage method

In this paper, a two-stage method for identification of the sandwich systems with hysteresis is proposed. In the first stage of this method, a monotonic and independently incremental signal is used to degenerate the hysteresis into a static function with one-to-one mapping. Thus, the sandwich system with hysteresis can be transformed into a system described by the combination of the transformed variables and the separated parameters based on the key term separation principle. Therefore, the parameters of the linear subsystems are estimated by the Recursive General Identification Algorithm (RGIA) since the transformed model is of the form containing linear coefficients. Then, in the second stage, based on the obtained linear subsystem models, the sandwich system with hysteresis is re-excited by a sine signal with attenuated amplitude. In this case, the hysteresis subsystem can be identified using neural networks based on the expanded input space method. Finally, the corresponding experimental results are presented.