陳明新臺灣大學：機械工程學研究所陳祈澈Chen, Chi-ChuChi-ChuChen2007-11-282018-06-282007-11-282018-06-282004http://ntur.lib.ntu.edu.tw//handle/246246/61163傳統的順滑模態控制只能處理滿足某個結構性條件-稱為匹配條件(matching condition)的不確定性統。本篇論文對於具有任意結構的非匹配不確定性系統提出了一個新的順滑模態控制方法。此新控制方法的獨特性在於建立了一個具有強韌特性的觀測器，藉由此觀測器，系統的外部狀態(external state)可以被估測得到。此觀測器同時使用了迴路轉移函數回歸法(Loop transfer recovery)和可變結構系統(Variable structure system)控制設計來確保漸進穩定的估測。當系統具有大量匹配與非匹配之不確定性時，此新的控制方法能夠達到漸近穩定之輸出追蹤。Conventional sliding mode control can only deal with system uncertainties that satisfy a structural condition called the matching condition. This thesis proposes a new sliding mode control for systems with arbitrarily structured mis-matched uncertainties. The unique feature of the new control is a robust observer for estimation of the external state of the system. The robust observer uses both the loop transfer recovery (LTR) design and variable structure system (VSS) design to ensure asymptotic estimation. The new control is capable of achieving asymptotic output tracking in the face of large matched or mis-match uncertainties.Abstract-Chinese version I Abstract-English version II Index of Contents III Index of Figures IV Index of Contents Chapter 1. Introduction 1 Chapter 2. Control for systems with matched uncertainties 3 Example 2.1 6 Example 2.2 7 Chapter 3. Systems with mis-matched uncertainties 8 3.1 VSS/LTR observer design 9 3.2 Cascade VSS/LTR observer design 11 Chapter 4. Control for systems with mis-matched uncertainties 13 Example 4.1.1 15 Example 4.1.2 15 Example 4.1.3 15 Example 4.1.4 16 Example 4.2.1 16 Example 4.2.2 16 Example 4.2.3 16 Example 4.2.4 17 Chapter 5. Conclusion 18 Reference 19 Figure 20 Index of Figures Figure 1 The LTR result for different dimensions of z 21 Figure 2. 1. 1 Time history of system state 22 Figure 2. 1. 2 Control input with time 22 Figure 2. 2. 1 Time history of system state 23 Figure 2. 2. 2 Control input with time 23 Figure 4. 1. 1. 1 Time history of system state 24 Figure 4. 1. 1. 2 Control input with time 24 Figure 4. 1. 1. 3 Estimation error of output derivatives with time 25 Figure 4. 1. 2. 1 Time history of system state 26 Figure 4. 1. 2. 2 Control input with time 26 Figure 4. 1. 2. 3 Estimation error of output derivatives with time 27 Figure 4. 1. 3. 1 Time history of system state 28 Figure 4. 1. 3. 2 Control input with time 28 Figure 4. 1. 3. 3 Estimation error of output derivatives with time 29 Figure 4. 1. 4. 1 Time history of system state 30 Figure 4. 1. 4. 2 Control input with time 30 Figure 4. 1. 4. 3 Estimation error of output derivatives with time 31 Figure 4. 1. 4. 1 Time history of system state 30 Figure 4. 1. 4. 2 Control input with time 30 Figure 4. 1. 4. 3 Estimation error of output derivatives with time 31 Figure 4. 2. 1. 1 Time history of system state 32 Figure 4. 2. 1. 2 Control input with time 32 Figure 4. 2. 1. 3 Comparison of output derivatives estimation error 33 Figure 4. 2. 2. 1 Time history of system state 34 Figure 4. 2. 2. 2 Control input with time 34 Figure 4. 2. 2. 3 Comparison of output derivatives estimation error 35 Figure 4. 2. 3. 1 Time history of system state 36 Figure 4. 2. 3. 2 Control input with time 36 Figure 4. 2. 3. 3 Comparison of output derivatives estimation error 37 Figure 4. 2. 4. 1 Time history of system state 38 Figure 4. 2. 4. 2 Control input with time 38 Figure 4. 2. 4. 3 Comparison of output derivatives estimation error 39483244 bytesapplication/pdfen-US順滑模態控制非匹配型不確定性可變結構系統輸出追蹤匹配條件matching conditionsliding mode controloutput trackingmis(un)-matched uncertaintyvariable structure systemthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/61163/1/ntu-93-R91522830-1.pdf