Abstract
For the purpose of understanding the physical and chemical properties of the nano material as well as performing processing work on it, a manipulator with accuracy in the nano meter range is absolutely necessary. Typically, these nano-scale manipulators are driven piezoelectric actuators. Regardless of the advantages such as lightweight, fast response and high resolution, the piezoelectric actuator also exhibits hysteresis behavior which will seriously limit its performance. In order to achieve more precise control on the piezoelectric actuator, this research is devoted to identifying the hysteresis and developing an algorithm to eliminate the positioning error caused by such an effect. Particularly, we utilize the Maxwell model and Bond-Graph representation to model hysteresis and use the experimental data to estimate the parameters in the model. The model acquired is then used to design a feedforward control so that the piezoelectric actuator can have better positioning performance even when controlled in the open-loop manner.