|
The so-called piezomechanics contains three parts: piezoelectric translator, carriage mechanism and control system. It is well-known that a piezoelectric translator consists of the following advantages: (i) unlimited resolution, (ii) no moving parts, (iii) high efficiency, (iv) large forces, and (v) fast response. However, three drawbacks of piezomechanics: (i) it should only be load axially, (ii) it contains hysteresis feature, and (iii) the expansion is dependent on temperature, must be tackled. The first drawback is solved by the carriage mechanism design. The thesis focuses on the secondand third drawbacks by using an intelligent variable structure control (IVSC).First, a neural-network, including two different nonlinear gains and a lineardynamic system, is employed to learn the dynamics of piezomechanism. Then a forward control based on this learned model is used to achieve an acceptabletracking result. Because the tracking performance using forward control cannot be guaranteed as the system is subject to uncertainties, a discrete-time variable structure control (DVSC) is then synthesized with the previous forward control to improve the tracking performance. No state estimator is required for the proposed control. The stability of the overall system is verified by the Lyapunov stability criterion and the experiment is also presented to confirmthe usefulness of the proposed control.
|