-
Notifications
You must be signed in to change notification settings - Fork 0
/
The Poincare-Friedrichs inequality on polygon and its applications.tex
571 lines (524 loc) · 21.9 KB
/
The Poincare-Friedrichs inequality on polygon and its applications.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
%The final version by LS on Oct 10th, 2016
%The manuscript by LS on Oct 8th, 2016
\documentclass[12pt]{article}
%%% The default page size is A4 piece of paper.
%%% In "[]", we can set the page size,also can set the font for your documents.
%%% \documentclass[11pt,a5paper]{article}
\usepackage{amsfonts}
\usepackage{amsmath}
%%%use (\usepackage[fleqn]{amsmath}) can change the formula that was center aligned into the left.
\usepackage{bm}
\usepackage{amssymb}
\usepackage{graphicx}
\usepackage{latexsym}
\usepackage{booktabs}
\usepackage{tabularx}
\usepackage{indentfirst}
%%%Features: Makes two blank space at the beginning of each paragraph.
\usepackage{cite}
\usepackage{mathrsfs}
\usepackage{caption2}
\usepackage{CJK}
\usepackage{stmaryrd}
\usepackage{subfigure}
\usepackage{overpic}
\usepackage{enumerate}
\usepackage{cases}
\usepackage{algorithm}
\usepackage{algorithmic}
\usepackage{ctex}
\usepackage{color,xcolor}
%%% added by Lin in the following
\usepackage{url} %%% add underlining on a mailbox by "\_"
\usepackage{slashbox}
\usepackage[active]{srcltx}
\usepackage{verbatim} %add comments H
\usepackage{float}
\usepackage{times} %%% Sen added:Uncomment to use the Times New Roman font
\usepackage{tikz} %%% figure
\usepackage{picinpar} %%%window,figwindow,tabwindow
%\DeclareMathOperator*{\argmin}{arg\,min}%% 自动底标了
%\synctex=0
%\usepackage{geometry} %设置页边距的宏包
%\usepackage{titlesec} %设置页眉页脚的宏包
%%% 如果想要自由设置页面属性的话,可以使用geometry宏包。
%%% geometry宏包可以让我们自由设置上下左右的页边距。
%%% \geometry{left=2cm,right=2cm,top=2.5cm,bottom=2.5cm} %设置 上、左、下、右 页边距
%%%%%%% added by Einstein Lin
%%%%% Definition of the environmental category theorem %%%%
%\providecommand{\U}[1]{\protect\rule{.1in}{.1in}} \textwidth 15.2cm
%\textheight 23.4cm \hoffset -1.5cm \voffset -2.2cm
%%%%\newtheorem{theorem}{\bf Theorem}[section]
%%%% the added '[section]' means theorem numbered by section
%\newtheorem{theorem}{\bf Theorem}
%\newtheorem{remark}{Remark}
%\newtheorem{lemma}{Lemma}
%\newtheorem{example}{Example}
%\newtheorem{defn}{Definition}
%\newtheorem{cor}{Corollary}
%\newtheorem{axiom}{Axiom}
%\newtheorem{proposition}{Proposition}
%\newtheorem{condition}{Condition}
%\newtheorem{conclusion}{Conclusion}
%
%\renewcommand{\figurename}{Figure}
%\renewcommand{\tablename}{Table}
%\renewcommand\refname{Reference}
%\floatname{algorithm}{Algorithm}
%
%\newenvironment{proof}[1][Proof]{\textbf{#1.} }
%{\ \rule{0.75em}{0.75em}\smallskip}
%%\newenvironment{proof}[1][\bf Proof]{\textbf{#1.} }{}
%\newenvironment{question}[1][\bf\large Que]{\noindent\textbf{#1:} }{\par}
%%\numberwithin{equation}{section} %%%number the formula by section
%\newcommand{\ucite}[1]{$^{\mbox{\scriptsize \cite{#1}}}$}
%\def\QEDopen{{\setlength{\fboxsep}{0pt}\setlength{\fboxrule}{0.2pt}\fbox{\rule[0pt]{0pt}{1.3ex}\rule[0pt]{1.3ex}{0pt}}}} %Define hollow characters
%\def\QED{\QEDopen}
%%% The following is for inserting the code in our Report
\usepackage{listings} % Insert codes
%%%for code settings
\lstset{language=Matlab}
\lstset{breaklines}
\lstset{extendedchars=false}
\lstset{ %
extendedchars=false, % Shutdown no-ASCII compatible
language=Matlab, % choose the language of the code
basicstyle=\footnotesize\tt, % the size of the fonts that are used for the code
tabsize=3, % sets default tabsize to 3 spaces
numbers=left, % where to put the line-numbers
%numberstyle=\tiny, % the size of the fonts that are used for the line-numbers
%stepnumber=1, % the step between two line-numbers. If it's 1 each line will be numbered
%numbersep=5pt, % how far the line-numbers are from the code
keywordstyle=\color[rgb]{0,0,1}, % keywords
commentstyle=\color[rgb]{0.133,0.545,0.133}, % comments
stringstyle=\color[rgb]{0.627,0.126,0.941}, % strings
backgroundcolor=\color{white}, % choose the background color. You must add \usepackage{color}
showspaces=false, % show spaces adding particular underscores
showstringspaces=false, % underline spaces within strings
showtabs=false, % show tabs within strings adding particular underscores
% frame=single, % adds a frame around the code, we have 2 Options: "shadowbox" & "single"
frame=shadowbox, %xleftmargin=2em,xrightmargin=2em, aboveskip=1em;
captionpos=b, % sets the caption-position to bottom
breaklines=true, % sets automatic line breaking
breakatwhitespace=false, % sets if automatic breaks should only happen at whitespace
title=\lstname, % show the filename of files included with \lstinputlisting;
% also try caption instead of title
mathescape=true,escapechar=? % escape to latex with ?..?
escapeinside={\%*}{*)}, % if you want to add a comment within your code
%columns=fixed, % nice spacing
%morestring=[m]', % strings
%morekeywords={%,...},% % if you want to add more keywords to the set
% break,case,catch,continue,elseif,else,end,for,function,global,%
% if,otherwise,persistent,return,switch,try,while,...},%
}
%----------------------------------------------------------------------------------------
% DOCUMENT INFORMATION
%----------------------------------------------------------------------------------------
\begin{document}
\begin{CJK*}{GBK}{song}
%%%\theoremstyle{break}
%%%\theorembodyfont{\kaishu}
%%%% 定理类环境的定义 %%%%
\renewcommand{\figurename}{图}
\renewcommand{\tablename}{表}
\newtheorem{theorem}{\bf 定理}
\newtheorem{remark}{\bf 注}
\newtheorem{lemma}{\bf 引理}
\newtheorem{assumption}{假设}
\floatname{algorithm}{算法}
\renewcommand\refname{参考文献}
\newtheorem{definition}{定义}
\newtheorem{axiom}{公理}
\newtheorem{proposition}{命题}
\newtheorem{corollary}{推论}
\newtheorem{condition}{条件}
\newtheorem{conclusion}{结论}
\newtheorem{property}{性质}
\newcommand{\ucite}[1]{$^{\mbox{\scriptsize \cite{#1}}}$}
\def\QEDopen{{\setlength{\fboxsep}{0pt}\setlength{\fboxrule}{0.2pt}\fbox{\rule[0pt]{0pt}{1.3ex}\rule[0pt]{1.3ex}{0pt}}}} % 定义空心符
\def\QED{\QEDopen}
\newenvironment{proof}[1][\bf 证明]{\textbf{#1.} }{\QED}%{}
\newenvironment{question}[1][\bf\large 问题]{\noindent\textbf{#1:} }%{?\par}
%\def\QED{\QEDopen}
%\newenvironment{solution}[1][\bf 解答]{\textbf{#1.} }{\QED}
\title{\emph{VEM Theory Analysis}\\\textbf{The Poincar\'{e}-Friedrichs inequality on polygon and its applications}}
% Title
\author{\Large{\emph{Submitted by:}}\\\\\Large{\textbf{JGH Seminars}}\\\textbf{School of Mathematical Science}\\\textbf{Shanghai Jiao Tong University in Shanghai}\\}
%\author{JGH Seminars\\{\small(SJTU, School of Mathematical Science, Shanghai, 200240)}}
% Author name
\date{}
\maketitle
\begin{center}
{\Large 多角形上的Poincar\'{e}-Friedrichs不等式及其应用}
\end{center}
%\noindent{\bf Key Words}: \qquad Polygon Division,Centroid Coordinate,Matlab code.
\vspace{1pc}
%\noindent{\bf Abstract}: This article mainly help students to understand the homeworks better.
%\vspace{1pc}
%%%\begin{figure}[b]
%%%\rule[-2.5truemm]{5cm}{0.1truemm}\\[2mm]{
%%%\quad\quad {\bf收稿日期:}\,2016-09-27\, \\
%%%{\bf 作者简介:}\,黄建国, 男, 教授, 博士生导师,\\
%%% \mbox{}\qquad\qquad~\,林森, 男, 博士研究生, 研究方向以有限元方法为主. \\
%%% \mbox{}\qquad\qquad~\,电话: 021-54743151-2105; \\
%%% %\mbox{}\qquad\qquad~\,E-mail:jghuang@sjtu.edu.cn \& linsenlyj@163.com }
%%% \mbox{}\qquad\qquad~\,E-mail:jghuang@sjtu.edu.cn \& sjtu\_Einsteinlin7@sjtu.edu.cn }
%%%\end{figure}
%%%\vspace{0.1mm}
%%\begin{figure}[b]
%%\rule[-2.5truemm]{5cm}{0.1truemm}\\[2mm]{
%%\quad {\bf Submitted date:}\,2016-10-08\, \\
%%{\bf Author:}\,Sen Lin, Male,PhD.\\
%% \mbox{}\qquad\quad~\,Research on the Finite Element Method(FEM). \\
%% \mbox{}\qquad\quad~\,Tel:13262292127; \\
%% \mbox{}\qquad\quad~\,E-mail:linsenlyj@163.com.}
%% %\mbox{}\qquad\qquad~\,E-mail:jghuang@sjtu.edu.cn \& sjtu\_Einsteinlin7@sjtu.edu.cn }
%%\end{figure}
%%\vspace{0.1mm}
%%% Generate directory
\newpage
\tableofcontents
\newpage
%\numberwithin{equation}{section} %number formula by section
%%% Section 1
\section{多角形的Poincar\'{e}-Friedrichs不等式}
由等价模定理可以推得
\begin{itemize}
\item Poincar\'{e}不等式:
\[
\|v\|_{0,K}\leq{C}\big(|v|_{1,K} + |\int_{K}v(x)\rm{d}x|\big)
\quad \forall v \in H^1(K);
\]
%%% \begin{align*}
%%% \|v\|_{0,K}\leq{C}\big(|v|_{1,K} + |\int_{K}v(x)\rm{d}x|\big)
%%% \quad \forall v \in H^1(K);
%%% \end{align*}
\item Friedrichs不等式:
\[
\|v\|_{0,K}\leq{C}\big(|v|_{1,K} + |\int_{\partial{K}}v(x)\rm{d}s(x)|\big)
\quad \forall v \in H^1(K).
\]
\end{itemize}
上述两式是$p=2$的结果,进一步,我们可以得到更一般的结果.
\begin{equation*}
\label{Eq:PFIeq}%%%Poincar\'{e}-Friedrichs inequality for W^{1,p}.
\left\{
\begin{aligned}
\|v\|_{L^p(K)} & \leq{C}\big(|v|_{W^{1,p}(K)}+|\int_{K}v(x)\rm{d}x|\big)
& \forall v \in W^{1,p}(K); \\
\|v\|_{L^p(K)} & \leq{C}\big(|v|_{W^{1,p}(K)}+|\int_{\partial{K}}v(x)\rm{d}s(x)|\big)
& \forall v \in W^{1,p}(K).
\end{aligned}
\right.
\end{equation*}
其中,$K\subset\mathbb{R}^2$是一个有界且连通的多角形区域和$1\leq{p}\leq\infty$.
\begin{itemize}
\item 1.由紧性论证技巧可证明:$C=C(K)$,即正常数$C$ 仅取决于$K$.
\item 2.若$K$为三角形,可由\textbf{scaling-argument} 来获得一致估计.
\end{itemize}
\begin{question}
\label{Q1}
当$K$为一般的多角形时,如何处理?即它需要满足什么要求时,可以得到一致估计?
\end{question}
下面,我们假设多角形$K$满足如下\textbf{性质P}:
\begin{property}[\large\textbf{P}] \label{Pro1}
有以下三点: \\
(1)存在$K$的正规三角剖分$\mathcal{T}_{h_K}(K)$,这里正规(shape-regularity),
意指,存在常数$C_r>0$使
\[
h_{\tau}/\rho_{\tau} \leq C_r < \infty, \quad \forall \tau \in \mathcal{T}_{h_K}(K),
\]
且$K$的每条边必为$\mathcal{T}_{h_K}(K)$中某一单元$\tau$之边; \\
(2)$\mathcal{T}_{h_K}(K)$中的单元个数是一致有界的,即$\#\mathcal{T}_{h_K}(K)\leq{L}$; \\
(3)在$K$中包含以半径为$r$的一个圆,且$h_K\lesssim{r}$, 其中生成的正常数$C$有界.
\end{property}
进一步,我们给出星形区域的定义,并给出两个例子来说明.
\begin{definition}[Star-shaped区域]\label{D1} \mbox{}\par
我们称区域$\Omega\subset\mathbb{R}^2$关于某个圆$B$(3 维为球)是星形的,
如果对于$\forall{x}\in\Omega$,$\{x\}\cup{B}$ 的闭凸包还是$\Omega$的子集.
\end{definition}
\begin{figure}[!h]
\centering
\subfigure[Domain star-shaped with respect to $B_1$ but not $B_2$]{
\label{Fig1:subfig:1} %% label for first subfigure
\includegraphics[width=2.1in,height=2.1in]{StarDomain01.eps}} %{StarDomain1.eps}
\hspace{0.05in}
\subfigure[Domain which is not star-shaped with respect to any circle $B$]{
\label{Fig1:subfig:2} %% label for second subfigure
\includegraphics[width=2.1in,height=2.1in]{StarDomain02.eps}}
\caption{星形区域的两个例子}
\label{Fig1} %% label for entire figure
\end{figure}
从图\ref{Fig1}可以看出图\ref{Fig1:subfig:1} 中的区域$\Omega$是星形区域,而
图\ref{Fig1:subfig:2}中的区域$\Omega$却不是星形区域.
\begin{proposition}\label{P1}
若多角形$K$为星形区域(star-shaped domain)且相应圆的半径$r$满足$h_K\lesssim{r}$,
且$\mathcal{T}_{h_K}(K)$中的每一单元$\tau$满足$h_{\tau}\approx{h_{K}}$,则
显然$K$满足\textbf{性质\ref{Pro1}}.
\end{proposition}
\begin{remark} \label{R1} %\mbox{}\par
在此,我们想说明一点:一方面,$h_{\tau}\lesssim{h_{K}}$ 显然成立;另一方面,由
\[
L*\pi(\frac{h_\tau}{2})\geq |K| \geq \pi{r}^2 \gtrsim\pi h_K^2
\]
可得 $h_{K}\lesssim{h_{\tau}}$. \\
正如下面的示意图所示,我们发现当多角形不满足性质\ref{Pro1}中(3)的条件
$h_K\lesssim{r}$时,由上述推导可知,我们无法得到$h_{\tau}\approx{h_{K}}$,进而无法
得到$|u-u_h|_{1,K}$的一致估计.这进一步说明了性质\ref{Pro1}(3)的重要性!
\begin{figure}[!ht]\label{Fig2}
\centering
%\includegraphics[width=3.2in,height=1.2in]{Diamond1.eps}
\includegraphics[width=3.6in,height=1.2in]{Diamond2.eps}
\caption{不满足性质\ref{Pro1}中的第3条的多角形示意图}
\end{figure}
\end{remark}
\begin{lemma}\label{L1}
给定一个三角形单元$\tau$,其满足$h_\tau/\rho_{\tau}\leq{C_r}<\infty$,设$e$为
$K$的任一边,则成立
\begin{equation}
\label{Eq:FIeq}
\|v\|_{0,p,K} \lesssim {h}_{\tau}^{-1}|v|_{1,p,K} + h_{\tau}^{2/p-1}|\int_ev\rm{d}s|
\quad \forall v\in W^{1,p}(K).
\end{equation}
式中$1\leq p<\infty$.
\end{lemma}
由\textbf{scaling argument}和参考元上的Friedrichs不等式立知结果.
\begin{theorem}[带几何尺度的Friedrichs不等式]
\label{T1}\mbox{}\par
假设多角形$K$满足\textbf{性质\ref{Pro1}},则对任何$v\in{W}^{1,p}(K)$
使$\int_{\partial{K}}v\rm{d}s=0$,成立估计
\begin{equation}
\label{Eq:GFIeq} %%% Friedrichs Inequality with Geometric scale
\|v\|_{L^p(K)} \lesssim {h_K}|v|_{1,p,K}
\end{equation}
式中生成常数与$h_K$无关.
\end{theorem}
\begin{proof}
\label{Proof1}
我们分以下几步给出证明. \\
\textbf{step 1.}设$h_K=diam(K)=1$,故由$h_{\tau}\approx{h_K}$知
$h_{\tau}\approx1$.设$\partial{K}$由$N$条边$e_1,e_2,\cdots,e_N$ 组成,
对任一边$e$,定义
\[
A_ev = \frac{1}{|e|}\int_ev\rm{d}s.
\]
则由条件$\int_{\partial{K}}v\rm{d}s=0$知
\[
\sum_{i=1}^N |e_i|A_{e_i}v = 0 \Rightarrow
\sum_{i=1}^N \frac{|e_i|}{|\partial{K}|} A_{e_i}v = 0.
\]
进一步,有
\begin{equation}
\label{Eq:GFIeq_LpEst1}
\begin{aligned}
\|v\|_{L^p(K)} & = \|v - \sum_{i=1}^N \frac{|e_i|}{|\partial{K}|} A_{e_i}v\|_{L^p(K)} \\
& = \|\sum_{i=1}^N \frac{|e_i|}{|\partial{K}|} (v - A_{e_i}v)\|_{L^p(K)} \\
& \leq \sum_{i=1}^N \frac{|e_i|}{|\partial{K}|} \|(v - A_{e_i}v)\|_{L^p(K)} \\
& \leq \max_{1\leq{i}\leq{N}} \|(v - A_{e_i}v)\|_{L^p(K)}.
\end{aligned}
\end{equation}
\textbf{step 2.} 记$w_i = v - A_{e_i}v$,$1\leq{i}\leq{N}$.由定义知
\begin{equation}
\label{Eq:GFIeq_LpEst2}
\|w_i\|_{L^p(K)} \leq \sum_{\tau\in\mathcal{T}_{h_K}(K)} \|w_i\|_{L^p(\tau)}
\leq {L} \|w_i\|_{L^p(\tau^{*})}
\end{equation}
式中
\[
\|w_i\|_{L^p(\tau^{*})} = \max_{\tau\in\mathcal{T}_{h_K}(K)} \|w_i\|_{L^p(\tau)}
\]
对于$\tau^{*}$可以找到一列三角形单元$\tau_1,\tau_2,\cdots,\tau_l$使得
$\tau_1=\tau^{*}$,$\tau_l=$以$e_i$为边的三角形(可参考下面示意图),其中$\tau_j$和
$\tau_{j+1}$的公共边为$f_j$,$1\leq{j}\leq{l-1}$,则由$h_{\tau_j}\approx 1$,引理\ref{L1}
和迹定理知
\begin{equation}
\label{Eq:GFIeq_LpEst3}
\begin{aligned}
\|w_i\|_{L^p(\tau^{*})}
& = \|w_i\|_{L^p(\tau_1)} \\
& \lesssim |w_i|_{1,p,\tau_1} + |\int_{f_1}w_i\rm{d}s|
& \text{(引理\ref{L1})} \\
& \lesssim |w_i|_{1,p,\tau_1} + \|w_i\|_{L^p(f_1)} \\
& \lesssim |w_i|_{1,p,\tau_1} + \|w_i\|_{1,p,\tau_2}
& \text{(迹定理)} \\
& \lesssim |w_i|_{1,p,\tau_1} + |w_i|_{1,p,\tau_2} + |\int_{f_2}w_i\rm{d}s|
& \text{($\tau_2$和$\tau_3$的公共边$f_2$)} \\
& \lesssim \sum_{j=1}^{l-1} |w_i|_{1,p,\tau_j} + \|w_i\|_{1,p,\tau_l}
& \text{(依此类推)} \\
& \lesssim \sum_{j=1}^{l} |w_i|_{1,p,\tau_j}
& \text{(引理\ref{L1} 和 $A_ev$ 定义)} \\
& \lesssim \big(\sum_{j=1}^{l} |w_i|_{1,p,\tau_j}^{p}\big)^{1/p}
\big(\sum_{j=1}^l 1\big)^{\frac{p-1}{p}}
& \text{(C-S不等式)} \\
& \lesssim l^{\frac{p-1}{p}} |w_i|_{1,p,K}
\end{aligned}
\end{equation}
由\eqref{Eq:GFIeq_LpEst1}-\eqref{Eq:GFIeq_LpEst3} 知
\begin{equation}
\label{Eq:GFIeq_1} %%% Friedrichs Inequality with Geometric scale one for h_K=1
\|v\|_{L^p(K)} \lesssim |v|_{1,p,K}
\end{equation}
\begin{figure}[!ht]\label{Fig3}
\centering
% \includegraphics[width=3.2in]{PolygonPartition1.eps}
\includegraphics[width=3.2in,height=3.2in]{PolygonPartition1.eps}
\caption{多角形$K$作正规剖分后得到一系列子三角形}
\end{figure}
%\begin{figure}[!h]\label{Fig2}
% \centering
% \includegraphics[width=3.2in]{PolygonPartition.eps}
% \caption{多角形$K$作正规剖分后得到一系列子三角形}
%\end{figure}
\textbf{step 3.} 对于一般情形,作变量代换$x=h_K\hat{x}$,$K\rightarrow\hat{K}$,
$diam(\hat{K})=1$,则由\eqref{Eq:GFIeq_1}立知结果\eqref{Eq:GFIeq}.
\end{proof}
使用与定理\ref{T1}中类似的推导可得
\begin{theorem}[带几何尺度的Poincar\'{e}不等式]
\label{T2}\mbox{}\par
假设多角形$K$满足\textbf{性质\ref{Pro1}},则对任何$v\in{W}^{1,p}(K)$ 使
$\int_{K}v\rm{d}s=0$,成立估计
\begin{equation}
\label{Eq:GPIeq} %%% Poincar\'{e} Inequality with Geometric scale
\|v\|_{L^p(K)} \lesssim {h}_K^{-1}|v|_{1,p,K}
\end{equation}
式中$1\leq{p}<\infty$,生成常数$C$与$h_K$无关.
\end{theorem}
%%% Section 2
\section{函数逼近的误差估计}
本节,我们主要给出关于函数逼近的误差估计的两个定理.
\begin{theorem}
\label{T3} %%%\mbox{}\par
假设多角形$K$满足\textbf{性质\ref{Pro1}},则对$\forall{v}\in{W}^{1,p}(K)$,
存在$v_{\pi}\in{P_k(K)}$使
\begin{equation}
\label{Eq:PAEst} %%% Polynomial Approximation Estimate
|v - v_{\pi}|_{l,p,K} \lesssim {h}_K^{k+1-l} |v|_{k+1,p,K},
\quad 0\leq{l}\leq{k}.
\end{equation}
\begin{proof}
\label{Proof2}
不妨设$h_K=1$证明结果即可(对于一般情形,作变量代换$x=h_K\hat{x}$). \\
找$v_{\pi}\in{P_k(K)}$使
\[
\int_K\partial^{\alpha}(v - v_{\pi}) \rm{d}x = 0, \quad |\alpha|\leq{k}.
\]
则由定理\ref{T2}知
\begin{align*}
\|v-v_{\pi}\|_{L^p(K)}
& \lesssim |v-v_{\pi}|_{1,p,K} \\
& \lesssim |v-v_{\pi}|_{2,p,K} \\
& \lesssim \cdots \\
& \lesssim |v-v_{\pi}|_{k+1,p,K} \\
& \lesssim |v|_{k+1,p,K}.
\end{align*}
\end{proof}
对于$\mathcal{T}_{h_K}(K)$,记
\[
\overline{V}_{h_K}(K) := \{ v\in{C}(\overline{K}): v|_e\in{P_k(e)},
\forall{e}\in\mathcal{T}_{h_K}(K) \}.
\]
$E_K$:$\overline{V}_{h_K}(K)$相应的$k$次Lagrange插值算子. \\
对$\forall{v}\in{H}^{k+1}(K)$,记$w=I_Kv$,由以下条件确定:
\begin{equation}
\label{Eq:IPDK} %%% Interpolation Projector definition on K
\left\{
\begin{aligned}
\Delta{w} & = \Delta{v_\pi} \in {P_{k-2}(K)}
& \text{in} \quad K; \\
w & = E_Kv\in{B}_k(\partial{K})
& \text{on} \quad \partial{K}.
\end{aligned}
\right.
\end{equation}
显然$w=I_Kv\in{V_k(K)}$. \eqref{Eq:IPDK}即
\begin{equation}
\label{Eq:IPDK1} %%% Interpolation Projector definition 1 on K
\left\{
\begin{aligned}
\Delta(w-v_\pi) & = 0
& \text{in} \quad K; \\
w-v_\pi & = E_Kv - v_\pi
& \text{on} \quad \partial{K}.
\end{aligned}
\right.
\end{equation}
由能量最小原理知
\[
|w-v_\pi|_{1,K} \leq |E_Kv - v_\pi|_{1,K}
\leq |E_Kv - w|_{1,K} + |w - v_\pi|_{1,K}
\]
进一步,有
\begin{equation}
\label{Eq:IPEsti_K} %%% Interpolation Projector Estimate on K
\begin{aligned}
|v-I_Kv|_{1,K} & = |v- w|_{1,K} \\
& \leq |v-v_\pi|_{1,K} + |w - v_\pi|_{1,K}
& \text{($\pm{v_\pi}$)} \\
& \leq |v-v_\pi|_{1,K} + |E_Kv - v_\pi|_{1,K} \\
& \leq 2|v-v_\pi|_{1,K} + |v - E_Kv|_{1,K}
& \text{($\pm{v}$)} \\
& \leq 2[|v-v_\pi|_{1,K} + |v - E_Kv|_{1,K}] \\
& \lesssim {h}_K^k |v|_{k+1,K}.
& \text{(\eqref{Eq:PAEst} 和$E_Kv$ 定义)}
\end{aligned}
\end{equation}
由定理\ref{T1}知
\begin{align*}
\|I_Kv - E_Kv\|_{0,K} & = \|w - E_Kv\|_{0,K} \\
& \lesssim {h}_K |w - E_Kv|_{1,K} \\
& \lesssim {h}_K^k |v|_{k+1,K}
\end{align*}
\begin{align*}
\|v - I_Kv\|_{0,K} & \leq \|v - E_Kv\|_{0,K} + \|I_Kv - E_Kv\|_{0,K}
& \text{($\pm{E_Kv}$)} \\
& \lesssim {h}_K^{k+1} |v|_{k+1,K}
\end{align*}
对$\forall{v}\in{H}^{k+1}(\Omega)$,定义:
\[
(I_hv)(x) := (I_Kv)(x) \quad x\in{K}.
\]
\end{theorem}
\begin{theorem}
\label{T4} %%%\mbox{}\par
假设多角形$K$满足\textbf{性质\ref{Pro1}},则对$\forall{v}\in{H}^{k+1}(K)$,
有$I_hv\in{V_h}$,且成立
\begin{equation}
\label{Eq:IPAEst} %%% Interpolation Projector Approximation Estimate
\left\{
\begin{aligned}
|v-I_hv|_{1,K} & \lesssim {h}_K^{k}|v|_{k+1,K}; \\
\|v-I_hv\|_{0,K} & \lesssim {h}_K^{k+1}|v|_{k+1,K}.
\end{aligned}
\right.
\end{equation}
\end{theorem}
由\eqref{Eq:IPDK}-\eqref{Eq:IPEsti_K},立即可得\eqref{Eq:IPAEst}.
%%% Section 3
\section{在VEM分析中的应用}
FEM: 找$u\in{V}:=H_0^1(\Omega)$使
\[
a(u,v) = f(v) \quad \forall {v}\in V.
\]
VEM: 找$u_h\in{V}_h$使
\[
a_h(u_h,v) = f_h(v) \quad \forall {v}\in V_h.
\]
式中$V_h\subset{V}$. \\
由第一Strang引理知
\begin{align*}
|u - u_h|_{1,\Omega}
& \lesssim |u - I_hu|_{1,\Omega} + \sup_{v\in{V_h}}|a(I_hu,v) - a_h(I_hu,v) |/|v|_{1,\Omega}
+ \|f - f_h\|_{-1,\Omega} \\
& \quad \text{(由$a(u_\pi,v)=a_h(u_\pi,v)$,其中分片$u_\pi\in\mathbb{P}_k$)} \\
% & \lesssim |u - I_hu|_{1,\Omega} + \sup_{v\in{V_h}}|a(I_hu-u_\pi,v) - a_h(I_hu-u_\pi,v)|/|v|_{1,\Omega}
% + \|f - f_h\|_{-1,\Omega} \\
& \lesssim |u - I_hu|_{1,\Omega} + |I_hu - u_\pi |_{1,\Omega} + \|f - f_h\|_{-1,\Omega} \\
& \quad \text{(由$|a(I_hu-u_\pi,v) - a_h(I_hu-u_\pi,v)|\leq{C}|I_hu - u_\pi|_{1,\Omega}|v|_{1,\Omega}$)} \\
& \lesssim |u - I_hu|_{1,\Omega} + |u - u_\pi|_{1,\Omega} + \|f - f_h\|_{-1,\Omega} \\
& \quad \text{(由$I_hu - u_\pi = (I_hu - u) + (u-u_\pi)$及三角不等式)} \\
& \lesssim h^k |u|_{k+1,\Omega} + h^k |u|_{k+1,\Omega} + h^k |f|_{k-1,\Omega} \\
& \quad \text{(由定理\ref{T4},定理\ref{T3}和$k\geq2$时)} \\
& \lesssim \mathcal{O}(h^k).
\end{align*}
\end{CJK*}
\end{document}