Hysteresis modeling of piezoelectric actuators using the fuzzy system

Abstract

An approach of hysteresis modeling in piezoelectric actuators is presented based on the multi-input single-output (MISO) fuzzy system. The proposed model adopts first-order Takagi-Sugeno (T-S) fuzzy system and transforms the multi-valued hysteresis into a one-to-one mapping with the extended input space vector. The generated fuzzy subspaces (multi-dimensional fuzzy sets) assign the maximum membership degree to the input data vectors. Fewer fuzzy subspaces are obtained by introducing the nearest neighbor and super radius concepts. The consequent parameter optimization is implemented after training the fuzzy system. Experimental results demonstrate that this methodology is algorithmically easy and can achieve high modeling accuracy. © 2010 Springer-Verlag.