顏家鈺Yen, Jia-Yush臺灣大學:機械工程學研究所葉雅琴Yeh, Yea-ChinYea-ChinYeh2010-06-302018-06-282010-06-302018-06-282009U0001-1708200912304800http://ntur.lib.ntu.edu.tw//handle/246246/187255壓電驅動的微米行程，奈米解析度的系統，在度量衡學與製程設備科技上，皆扮演很重要的角色。本論文提出兩種壓電驅動的光學掃描系統，應用於不同需求的近代光學系統上；一掃描系統主要在提供高解析與高扭力的步進掃描動作，另一掃描系統則可提供大範圍快速反覆的掃描動作。在光學掃描系統中，常需要作連續性的週期運動，本論文首先發展一種切換式不重置反覆學習控制器，用來改善週期性反覆的穩態誤差。切換控制學習的方法可以限制住在實現傳統反覆學習控制器時，常發生的不收斂重置誤差並減輕因變動的起始狀態所造成的不穩定性，以獲得週期性疊代收斂的穩態誤差。針對某些連續的週期性參考目標，切換式不重置反覆學習控制器可以將穩態誤差收斂到感應器的雜訊等級(零穩態誤差)。外，本論文也提出一個應用於電子束微影系統的多自由度的奈米定位載台與追蹤變結構控制器來解決其遲滯模型的不確定性。來自遲滯模型的系統不確定性可以被歸類成匹配式與非匹配式的系統不確定性。變結構控制器可以成功地抑制因為不同操作頻率，所產生的遲滯效應並維持其固定的伺服控制精度。PID控制器的伺服控制精度約為75奈米，而變結構控制器的定位伺服控制精度則可達40奈米，相較於使用PID控制器，變結構控制器展現40%的改善量。Piezo-actuated micro range and nano resolution systems play important roles in metrology and process equipment technology. Two kinds of piezo-actuated optical scanning systems for modern optics are proposed in this thesis. Two scanning systems differ in its scanning capability, one requires of high resolution step-and-scan motion and the other achieves wide range and fast scan motion. This thesis proposes a modification of the learning controller called “No-reset Switching Iterative Learning Control” to deal with the repetitive error of the optical scanning systems. This method provides a constraint to the non-converging reset error experienced by most conventional iterative learning control algorithms. Its ability to alleviate the influence of the changing initial states was also illustrated. Both the simulation results and the experimental results confirm the performance of the proposed control. For some piecewise continuous references, the steady state errors of the proposed method converge to the sensor noise range (zero steady state error).lso, a piezo-actuated multi-axis nano-positioning stage for SEM based E-beam lithography system are proposed in this thesis. A tracking variable structure control with hysteresis models is employed to solve hysteresis uncertainty. The system uncertainty from hysteresis model enters the system in the forms of both matched and unmatched uncertainties. The thesis shows that the variable structure control can successfully suppress frequency-dependent hysteresis while maintaining constant servo accuracy. The control accuracy of the VSC is within 40 nm, which is greater than 40% improvement over the 75 nm accuracy maintained by the PID controller.Abstract…………………………………………………………….………………………………………………………Ⅰable of Contents……………………………………………...………………………………………………………Ⅴist of Figures……………………………………………………………..……………………………………………Ⅸist of Tables………………………………………………………………………….……………………………..…XVHAPTER 1 INTRODUCTION 1.1 THESIS OUTLINE 1.2 CONTRIBUTIONS 2HAPTER 2 PIEZOELECTRIC PRECISION POSITIONING SYSTEMS 5.1 MOTIVATIONS AND LITERATURE REVIEWS 5.2 HIGH RESOLUTION TIP-AND-TILT PIEZOELECTRIC SCANNING PLATFORM 10.2.1 Design Procedures of the Tip-and-tilt Piezoelectric Scanning Platform 12.2.2 System architecture of tip-and-tilt piezoelectric scanning platform 19.3 WIDE RANGE NO-BEARING ROTATING PIEZOELECTRIC SCANNING PLATFORM 21.3.1 Design Procedures of Wide Range No-bearing Rotating Piezoelectric Scanning Platform 24.3.2 System architecture of no-bearing rotating piezoelectric scanning platform 34.4 MULTI-AXIS NANO-POSITIONING STAGE FOR THE E-BEAM LITHOGRAPHY SYSTEM 36.4.1 Mechanical design concept of the multi-axis nano-positioning Stage 38.4.2 System architecture of multi-axis nano-positioning stage 41.5 HYSTERESIS MODELS OF PIEZOELECTRIC ACTUATOR 42.5.1 Approximated Polynomial Model 42.5.2 Maxwell Slip Model 43.5.3 Duhem Model 45.5.4 Bouc-Wen Model 45.5.5 Preisach Model 46.5.6 TF model 47.6 MULTILAYER PIEZOELECTRIC ACTUATOR 51.6.1 Piezo-Stack－Piezomechanik：Pst 150/2×3/20 & Pst 150/5×5/20 53.6.2 Piezo Translator －PI：P-840.40 54.6.3 Amplified Piezo Actuator－Cedrat Technologies：APS40SM 56.7 PIEZO AMPLIFIER 57.7.1 Bipolar analog amplifier －Piezomechanik：SVR 150bip/3 58.7.2 LVPZT Piezo Amplifier－PI E-663.00 & PI E-530.00 60.8 LVDT 63.8.1 Principle of LVDT 64.8.2 LVDT (Solartron AGZ 5.0) & LVDT control module (PI E-509.L3) 66.8.3 LVDT’s Calibration 70.9 PSD 71.9.1 Principle of quadrant PSD 72.9.2 PSD calculating circuit 73.9.3 PSD calibration 74.10 DATA ACQUISITION CARD 76.10.1 Analog to digital transformation card 77.10.2 Digital to analog transformation card 77.10.3 Minimum detectable input voltage 78.11 HARDWARE LIMITATION OF PC BASED CONTROL SYSTEM 80HAPTER 3 SYSTEM IDENTIFICATION AND HYSTERESIS MODEL 85.1 LINEAR TIME INVARIANT PROPERTY AND SUPERPOSITION CHARACTERISTIC 86.1.1 Linear time invariant property test 87.1.2 Linear superposition relations between sub-TF 90.2 IDENTIFICATION METHODS 92.2.1 Kinematic model 93.2.2 State space dynamic model－parameter identification 95.2.3 Frequency domain dynamic model－instrument identification (Agilent 35670A) 101.3 TRANSFER FUNCTION HYSTERESIS MODEL 109.3.1 TF models for X-Y-θZ stage 110HAPTER 4 LEARNING CONTROL 113.1 ITERATIVE LEARNING CONTROL 113.1.1 Introduction 115.1.2 SISO first order ILC 117.1.3 MIMO first order ILC 122.2 NO-RESET SWITCHING ITERATIVE LEARNING CONTROL 125.2.1 Introduction 126.2.2 Design concept of no-reset iterative learning control 130.2.3 Design concept of no-reset switching iterative learning control 133.3 REPETITIVE CONTROL 137.3.1 Introduction 137.3.2 Design concept of repetitive control 139.3.3 Design concept of modified repetitive control 145HAPTER 5 VARIABLE STRUCTURE CONTROL 151.1 INTRODUCTION 151.2 VARIABLE STRUCTURE CONTROL ALGORITHM 151.2.1 The matched disturbance 155.2.2 Unmatched disturbance 156HAPTER 6 SIMULATIONS AND EXPERIMENTS 159.1 SIMULATIONS AND EXPERIMENTS OF LEARNING CONTROLS 159.1.1 Simulation of SISO CILC on tip-and-tilt piezoelectric scanning platform 160.1.2 Simulation of SISO NRSILC on tip-and-tilt piezoelectric scanning platform 162.1.3 The verifications of the NRSILC method from other servo mechanisms 164.1.4 Simulation and experiments of MIMO NRSILC on tip-and-tilt piezoelectric scanning platform 167.1.5 Simulation and experiments of SISO NRSILC on no-bearing rotating piezoelectric scanning platform 172.1.6 Simulation and experiments of SISO RC on no-bearing rotating piezoelectric scanning platform 175.2 SIMULATIONS AND EXPERIMENTS OF VSC ON MULTI-AXIS NANO-POSITIONING STAGE 178.2.1 Simulations and experiments of VSC on X-Y-θZ piezo-stage 178HAPTER 7 DISCUSSIONS & CONCLUSIONS 183.1 DISCUSSIONS AND CONCLUSIONS OF LEARNING CONTROL 183.2 DISCUSSIONS AND CONCLUSIONS OF ROBUST CONTROL 188EFERENCE： 1917179058 bytesapplication/pdfen-US壓電奈米定位系統反覆學習控制器切換式不重置反覆學習控制器追蹤變結構控制器piezoelectric nano-positioning systemiterative learning controlno-reset switching iterative learning controltracking variable structure controlthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/187255/1/ntu-98-F91522812-1.pdf