Modeling Inverse-Hysteretic Systems Based on Expanded Input Space

In order to improve the performance of the system with piezoelectric actuators, one of the approaches is to construct an inverse model of the hysteresis to cascade with the actuator so as to compensate for the effect of hysteresis involved in the piezoelectric actuators. In this paper, a neural-network-based inverse model for the hysteresis is proposed. In this scheme, an inverse hysteretic operator is proposed to extract the change tendency of the hysteresis inverse. Thus, an expanded input space that involves the inverse hysteretic operator as well as the input of the inverse hysteresis is constructed. This expanded input space is able to transform the multi-valued mapping of the inverse hysteresis into a kind of one-to-one mapping so that the neural networks are capable of implementing identification for the hysteresis inverse.