反覆式學習控制(Iterative Learning Control, ILC),是一經由反覆的控制系統過程中記錄系統所輸入、輸出的資訊,以作為下次控制力修正的參考,且反覆學習控制能達到在有限的學習次數中,將系統的跟隨誤差收斂到一有限的範圍內。但在學習過程中可能會受到外界干擾,導致學習效能不佳,因此利用小波轉換處理不可學習之動態。本文提出以小波轉換為基礎之無刷主軸馬達變轉速控制器,以解決傳統控制器難以處理的非線性以及不確定性的問題,使主軸馬達能追循命定軌跡的速度命令曲線。 實驗硬體架構方面,本文以光碟機主軸馬達為驅動系統之控制對象。但原本馬達回授之轉速訊號解析度過低,因此需利用一數位位置轉換器,對回授之轉速訊號加以處理,以提升回授訊號之解析度,並將搭配以小波轉換為基礎之控制器,以及速度估測器,使光碟機主軸馬達能追尋速度命令曲線。
The Iterative Learning Control (ILC) scheme is using the information in repetitive operation and progressive improving the tracking performance. The tracking error can be convergence to small range in a finite learning time. The performance of the iterative learning controller will be limited caused by non-repeatable disturbances. Thus wavelet transform was applied to handle the un-learnable dynamics. This thesis presents a wavelet-based controller on the brushless spindle velocity tracking. It can solve the nonlinear and uncertain problems that the traditional controller can not solve. Thus the spindle can track an arbitrary speed profile. About the hardware structure, the plant under study in the thesis is the spindle of an optical disk drive. Since the resolution of the speed signal from the spindle is too low, we choose a digital position converter (DPC) to improve the resolution. Thus that will enhance the performance of proposed controller.