| Summary: | 碩士 === 國立東華大學 === 電機工程學系 === 92 ===
In this thesis, the improvement of the tracking performance of piezoelectric actuators (PZA) for precision motion control is proposed. It is well known that the tracking performance of PZA is always deteriorated due to the nonlinear hysteresis and materials. Therefore, the neural network and fuzzy neural network controllers are developed, in which both the tracking controller and identifier are included in system control design. The neural controller is developed for micro-positioning control with high performance and the neural identification mechanism is simultaneously presented to provide the sensitivity information required by neural controller. In these developed controllers, analyses of the asymptotical stability for neural controller and identifier are presented. Finally, some experimental results are demonstrated to validate the proposed control designs for practical applications.
|
|---|
