• 沒有找到結果。

拋物線型分散式系統的控制器設計

N/A
N/A
Protected

Academic year: 2021

Share "拋物線型分散式系統的控制器設計"

Copied!
4
0
0

加載中.... (立即查看全文)

全文

(1)

1

行政院國家科學委員會專題研究計畫成果報告

拋物線型分散式系統的控制器設計

Contr oller Design for Par abolic Type

Distr ibuted Par ameter Systems

計畫編號:NSC 87-2218-E-002-029

執行期限:86 年 8 月 1 日至 87 年 7 月 31 日

主持人:馮蟻剛 臺灣大學電機系

一、中文摘要 在本計畫中,針對特定形式之拋物線 型分佈參數系統,研究一有限維控制器的 設計方式。 透過關於分佈參數系統的有限維近 似,可以最佳控制理論容易地設計有限維 的控制器,在本研究中,提出運用此有限 維控制器時,全系統穩定性及表現度的分 析方式。其結果則顯示,對於特定形式的 拋物線型分佈參數系統,在有限維的近似 模型下所設計的有限維控制器,是可以指 數化地穩定全系統,我們並可對其相對於 有限維近似模型的表現度變化進行估計。 關鍵詞:分佈參數系統,有限維控制器設 計,最佳控制理論 Abstr act

This research deals with the finite dimensional controller design problem for certain class of parabolic type distributed parameter systems. Based on a finite dimensional approximation model of the distributed parameter systems, we can design finite dimensional controllers via the optimal

control theory for finite dimensional systems. For these controllers, we make the stability and performance analysis of the overall system including the controller.

The result shows that for the class of

distributed parameter systems under

discussion, it is possible to design finite dimensional controllers directly based on the finite dimensional approximation model which exponentially stabilize the overall systems. We can also estimate the overall system performance change with respect to that of the finite dimensional model.

Keywor ds: distributed parameter systems, finite dimensional controller design, optimal control theory 二、緣由與目的 拋物線型分佈參數系統主要涵蓋由熱 傳導模型所描述的熱傳導系統。在分佈參 數系統的控制器設計上,若直接比照有限 維系統的控制器設計方式,則所設計出的 控制器亦常是一分佈參數系統。但在實際 工程應用上,由分佈參數系統所描述的控

(2)

2 制器,則無法進行實作。故而,對分佈參 數系統而言,有限維的控制器設計是在工 程應用上不可避免的要求。 在文獻上,關於有限維控制器的設 計,有著相當的成果[4,6,8,9]。其方法大致 分兩大方向:其一為直接設計出分佈參數 形式之控制器,再對此控制器做有限維之 近似;另一方式則是先求得分佈參數系統 的有限維近似模型,再依此近似模型求得 有限維的控制器。但如何保證全系統的閉 迴路穩定及其他相關特性,則是不可避免 的難題。 在本計畫的研究中,採用系統的有限 維近似模型來處理控制器的設計問題。經 由分佈參數系統的有限維近似模型,可以 利用在有限維系統中既有之最佳控制理 論,而推導最佳之控制器。在本研究中, 針對此控制器應用在閉迴路全系統後,全 系統之穩定性及表現度進行討論。並提出 穩定性及表現度所受到影響之估計。 對原分佈參數系統而言,本研究所提 出之方法對全系統而言並非是一最佳之控 制器設計,而是次佳(Sub-Optimal)之控制 器設計方式,但此設計方式卻是在現有的 計算環境中即可容易達成,而無須對複雜 的算子方程(Operator Equation)求解。 三、研究方法與成果    令 Z 為一希伯特空間(Hilbert Space), Z Z A D A: ( )⊂ → 為一定義在 Z 上的無界 算子(Unbounded Operator),D(A)在 Z 中為 稠 密 (Dense) , 且 A 為 封 閉 算 子 (Close Operator) 。 由 A 產 生 出 的 線 性 半 群 (C -0 Semigroup)為T0(t)[5]。考慮如下之系統        = = + = 0 ) 0 ( ) ( ) ( ) ( ) ( ) ( z z t Cz t y t Bu t Az t z dt d (1) 其 中 BL(U,Z),CL(Z,Y) 為 有 界 算 子 (Bounded Operator),B=[b1(x),...,br(x)], )] ( ),..., ( [u1 t u t uT = r假設 A為一自伴(Self Adjoint)[3]算子 且有緻密之預解集(Compact Resolvent)。令 i λ 與φ 滿足ij Aφij =λiφijj=1...mi則 A之 特徵向量φ 形成 Z 之一組基底。取ij 0 ,n> l , nl, 為正整數。 首先,定義出關於(1)式的有限維部分 狀態:令

∑∑

= = > < = n i m j ij ij n i z z P 1 1 ,φ φ n n I P Q = − 則由[7]可知,若 > =< ij ij z z ,φ ] , ,..., , [< 1 > < > = ij r ij ij b b b φ φ ] ,..., [ ) ( 1 i im i T i t z z z = ] ,..., [ ~ 1 imi i T i b b B = 則zi(t)滿足 ) ( ~ ) (t Bu t z z&i =λi i + ix1T =[z1(t),...,zl(t)],B1T =[B~1(t),...,B~l(t)] 及 [ ,..., ] 1 1 1 diag Im lIml A = λ λ ,則x1(t)滿足 ) ( ) ( 1 1 1 1 Ax t Bu t x&= + (2) 此外,令 )] ( ),..., ( [ 1 2 z t z t xT = l+ n )] ( ~ ),..., ( ~ [ 1 2 B t B t BT = l+ n ] ,..., [ 1 1 2 diag l Iml nImn A = λ+ + λx2(t)滿足 ) ( ) ( 2 2 2 2 Ax t B u t x& = + (3) 並假設rank(B~i)=mi使得(2),(3)描述之系 統具可控性。在輸出部分,若

(3)

3 ] ), ( ,..., ), ( [< 1 > < 1 > = c x z c x z yT 則令 > =< ij k k ij c x c ( ),φ           = p im p i im i i i i c c c c C Λ Μ Μ Λ 1 1 1 1 ~ ] ~ ,..., ~ [ 1 1 C Cl C = ] ~ ,..., ~ [ 1 2 Cl Cn C = + T p n n nu Q c u Q c u S =[< 1, >,...,< , >] 其中uQn(Z),則輸出方程式可表為: ) ( ) ( ) ( ) (t C1x1 t C2x2 t S Q zt y = + + n n 並假設rank(C~i)=mi,使得(2),(3)描述之 系統具可觀察性。 在控制器的設計上,採用狀態估測器 的迴授控制,定義狀態估測器為    + = − − + + = u B z A t z z C z C y G u B z A t z 2 2 2 2 2 2 1 1 1 1 1 1 1 ) ( ) ( ) ( & & (4) 且迴授控制為u=F1z1(t),其中F 及1 G 為1 待給定之矩陣。 Lemma 1 [7] 若F 及1 G 使得1 λ(A1+B1F1) 1 + <λ ,l λ(A1−G1C1)<λl+1,maxλ(A1 + 1 1 1F )> n+ B λ ,maxλ(A1G1C1) >λ ,且n+1 存在一σ 使得0>σ >λl+1,則由(2),(3)及 (4)所描述之有限維控制器在閉迴路下使全 系統為指數穩定。 根據 Lemma 1 的結果,我們提出一次 佳化(Sub-Optimal)的有限維控制器的設計 過程。首先考慮(2),其為關於系統(1)的有 限維近似系統,由最佳控制之理論[1],令 P 為 0 1 1 1 1 1 + − + = − B P Q R B P P A A P T T 之解,再令 ( ) 1( ) * t x K t u = ,其中 P B R K 1 1 * =− − 則受此控制後(2)之閉迴路系統將最小化如 下之表現指標:

∞ + = 0 1 1 2 ) (u Ru x Qx dt V T T (5) 令 * V 為此最小表現指標,即

∞ + = 0 1 * * 1 2 * ) (Q K RK xdt x V T T 但是對原系統(1)而言,以 * K 為控制器 是否能保證使閉迴路的全系統達到指數穩 定及最小化(5)之表現指標,則需加以討 論。 令L=m1 +....+mlM =ml+1+....+mn Z Q R R Z = LMn ,並令 ] , , [ 1 1 1 1 1 1 0 diag A BF A GC AQn A = + −           = 0 0 0 0 0 0 1 1 1 1 1 BF Q S G F B A n nA 為 定 義 在 Z 上 的 無 界 算 子0 (Unbounded Operator)。設T 為由0 A 所產生0 之線性半群,T 為由1 (A0 + A1)所產生之線 性半群。令 ) ( Re max A1+B1K* = λ ωω <0,且有µ >0使得 Tµewt 0 。首 先,考慮由 * K 所形成之控制器之穩定性: Theor em 1 令 F1 =K* , 選 取 G 使 得1 0 ) ( 11 1 <σ < λ A GC ,maxReλ(A1G1C1) 1 + >λ ,適當選取 n 之大小,則由(2),(3)n 及(4)所描述之有限維控制器在閉迴路下使 全系統為指數穩定。 若V~為全系統閉迴路時之表現指標, 即

(4)

4

∞ + = 0 1 * * 1 2 * ) ( ~ dt x RK K Q x V T T 其中x 為全系統閉迴路下之部分狀態。則1 在表現指標上有以下之結果: Theor em 2 令ω =λl+1+ωµ =µ A1µ λ λ ω ω µ µ 1 1 2 0 2 ) 2 )( ( ( + + + − − + = z k l k l p )) 1 (µ + , q=2λl+1ωω(µ +ω)(ω +µ) ) 2 (µ + ϖ ,則 q p V V* − ~ ≤ 。 四、結論與討論 在本報告中,對於一特定形式之拋物 線型分佈參數系統提出了一有限維控制器 的設計及分析方式。在設計方面,由現有 之控制計算工具即可計算出所需之控制 器;在分析方面,則提出在此有限維控制 器下全系統的穩定性分析及表現度估計。 五、計畫成果自評 針對原提之計畫,本計畫研究人員已 達成主要目的,提出一實際可行之控制器 設計即分析方式,。本計畫本質上即屬理 論性,因此所探討的問題較為一般化,所 取得之成果亦可應用於多種工程系統,例 如熱傳導系統的分析上。目前本計畫之研 究成果已在整理中,預計將投稿國際性學 術期刊,以期獲得更多支回饋而做進一步 之突破。 六、參考文獻

[1]Anderson, B. D. O., and J. B. Moore,

Optimal Control- Linear Quadratic Methods. Prentice Hall, New Jersey,

1990.

[2]Bensoussan, A., G. D. Prato, M. C.

Delfour, and S. K. Mitter, Representation

and Control of Infinite Dimensional Systems, Volume I, Birkhäuser, Boston,

1992.

[3]Conway, J. B., A Course in Functional

Analysis, Springer-Verlag, New York,

1990.

[4]Curtain, R. F., “A Comparison of

Finite-Dimensional Compensator Design for Distributed Parameter Systems,” Control - Theory and Advanced Technology, vol.

9, no. 3, pp. 609-628, 1993.

[5]Curtain, R. F. and H. Zwart, An

Introduction to Infinite-Dimensional Linear System Theory, Springer-Verlag,

New York, 1995.

[6]Li, Xunjing, and J. Yong, Optimal

Control Theory for Infinite Dimensional Systems, Birkhäuser, Boston, 1995.

[7]Pazy, A., Semigroups of Linear

Operators and Applications to Partial Differential Equations, Springer-Verlag,

New York, 1983.

[8]Sakawa, Y.,”Feedback Stabilization of

Linear Diffusion Systems,” SIAM J.

Control and Optim., vol. 21, no. 5, pp.

667-676., 1983.

[9]Schumacher, J. M., “A Direct Approach to Compensator Design for Distributed

Parameter Systems,” SIAM J. Control

and Optim., vol. 21, no. 6, pp. 823-836,

參考文獻

相關文件

圖4 1 整合資訊系統風險 圖4.1 整合資訊系統風險..

The finite difference equations will now be applied to solve the problem of a doubly drained clay layer, undergoing one dimensional consolidation. The assumptions made in deriving

• Learn the mapping between input data and the corresponding points the low dimensional manifold using mixture of factor analyzers. • Learn a dynamical model based on the points on

• Delta hedge is based on the first-order approximation to changes in the derivative price, ∆f , due to changes in the stock price, ∆S.. • When ∆S is not small, the

Based on Biot’s three-dimensional consolidation theory of porous media, analytical solutions of the transient thermo-consolidation deformation due to a point heat source buried in

Based on Biot’s three-dimensional consolidation theory of porous media, analytical solutions of the transient thermo-consolidation deformation due to a point heat source buried

In order to investigate the bone conduction phenomena of hearing, the finite element model of mastoid, temporal bone and skull of the patient is created.. The 3D geometric model

Wada H., Koike T., Kobayashi T., “Three-dimensional finite-element method (FEM) analysis of the human middle ear,” In: Hüttenbrink KB (ed) Middle ear mechanics in research