基於遲滯估測器與遲滯模型之混合式壓電平台遲滯補償

Abstract

由於具備高驅動解析度與高操作頻寬，壓電材料被廣泛應用於微控制領域之中，然其受制於遲滯效應－一種非線性之能量耗損－不僅驅動效率因而降低，控制成果也深受阻踞。 依據文獻中提出的電荷控制架構，藉由串連一個額外電容，可以準確地識別壓電制動器之等效模型，進而以電容上的壓降來估測遲滯所消耗的電壓。此外，此作法將控制目標由輸出位移轉換為電容上之電荷存量，不僅避免位置量測時之雜訊干擾，還能大幅降低控制成本。 基於上述硬體架構，本論文提出兩種補償方式來減少遲滯之影響，這兩種做法都是以遲滯觀察器來做為補償基礎，然而其中一個與遲滯模型進行結合，而另一個則是搭配內迴路控制器。這兩個補償機制只作用於遲滯消除而不進行任何的位置控制，因此在設計上位置控制器與遲滯補償兩者各自獨立。 除了補償方法所具備之運算架構，其所對應的穩定準則與效能分析也將在本論文內進行討論，並且在最後佐以不同控制頻率下之驗證結果。

With large bandwidth and high resolution, piezo-actuated stage is widely used in nanoscale position systems. However, the piezo is suffered from the hysteresis: a nonlinear effect which causes additional energy consumption that further degrades the control performance. According to the charge control structure proposed by L.S. Chen et al., by connecting a capacitor to the system, the parameter of the piezo actuator can be identified and further used to estimate the voltage consumed by the hysteresis. Based on the hardware structure described in the previous paragraph, this thesis proposes two compensation methods to eliminate the hysteresis effect. Both methods use a hysteresis observer to calculate the compensation value, however, only one of them is integrated with a hysteresis model, while the other is combined with an inner-loop controller. The two methods are only used for hysteresis cancellation and not involved in position control. Therefore, in the design process, the position controller and the hysteresis compensator are independent of each other. In addition to the compensation structure of the proposed methods, the corresponding stability criteria and performance analysis are also discussed in this thesis. Moreover, for validity, a series of experiments are also implemented, whereby the control results are presented in the end of this thesis.