summaryrefslogtreecommitdiff
path: root/macros/latex/contrib/fancytooltips/examples/fancy-preview-demo2.tex
blob: a756a44090cd0451bb333b04910353f4f99bcc65 (plain)
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
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
\documentclass{article}

\let\rmdefault\sfdefault
\def\modra#1{{\color{blue}\bm{#1}}}
\def\cervena#1{{\color{red}\bm{#1}}}
\def\separuj{\par\smallskip\hrule\kern 0.5pt\hrule \smallskip}
\def\separujB{\par\hrule\kern 0.5pt\hrule}

\newenvironment{block}{}{}
\usepackage{amsfonts,amsmath,amsthm,url,bm}
\usepackage{fancybox}
\usepackage{mathpazo}
\usepackage[latin2]{inputenc}
\usepackage[IL2]{fontenc}


\newtheorem{theorem}{Theorem}
\newtheorem{corollary}{Corollary}
\newtheorem{lemma}{Lemma}
\newtheorem{Theorem}{Theorem}
\def\theTheorem{\Alph{Theorem}}
\theoremstyle{definition}
\newtheorem{definition}{Definition}
\newtheorem{remark}{Remark}

\sloppy
\everymath{\displaystyle}
\usepackage[pdftex,nodirectory]{web}
\def\titlepageTrailer{}
\margins{.15in}{.15in}{12pt}{.15in} % left,right,top, bottom
\screensize{4.5in}{6in} % web.sty dimensions

\parindent 0 pt

\usepackage{mdwlist}
\usepackage{eso-pic}
\definecolor{mygreen}{RGB}{120,190,20}
\definecolor{mygreen}{RGB}{10,80,40}
\definecolor{webgreen}{RGB}{10,80,40}
\definecolor{seda}{gray}{0.31}
\definecolor{webgreen}{RGB}{120,190,20}
\AddToShipoutPicture{\hbox to 0 pt{\hbox to \paperwidth{\color{mygreen}\vrule
width 0.5em height\paperheight\color{black}%\hskip -0.5 em 
\hskip 0 pt plus 1 fill
\raise 1 pt\hbox {\normalfont\tiny \color{gray}\textbf{CDDEA 2010, Rajecké Teplice} (\thepage/12)} 
\hskip 0 pt plus 1 fill
}}}%

\def\qed{}

\def\lambdamin{\lambda_{\text{\rm{min}}}}
\def\lambdamax{\lambda_{\text{\rm{max}}}}

\makeatletter\let\over\@@over\makeatother
\def\theenumi{\roman{enumi}}
\def\labelenumi{\textrm{\upshape{(\theenumi)}}}
\def\konst{\textrm{const}}
\def\div{\mathop{\hbox{\rm div}}}
\def\meas{\mathop{\hbox{\rm meas}}}
\def\sgn{\mathop{\hbox{\rm sgn}}}
\def\laplac{\Delta}
\def\R{\mathbb{R}}
\def\N{\mathbb{N}}
\def\dxi{\,\mathrm{d}\xi\,}
\def\dx{\,\mathrm{d}x\,}
\def\dS{\,\mathrm{d}\sigma\,}
\def\dt{\,\mathrm{d}t\,}
\def\dT{\,\mathrm{d}T\,}
\def\du{\,\mathrm{d}u\,}
\def\ds{\,\mathrm{d}s\,}
\def\dr{\,\mathrm{d}r\,}
\def\dphi{\,\mathrm{d}\phi\,}
\newcommand{\duxi}{\frac{\partial u}{\partial x_i}}
\newcommand{\derxi}{{\partial\over\partial x_i}}
\newcommand{\pnorm}[1]{\|#1\|_p }
\newcommand{\qnorm}[1]{\|#1\|_q }
\newcommand\diver{\mathop{\rm div}}
\let\hat\widehat
\let\tilde\widetilde
\let\~\tilde

\let\phi\varphi

\def\vyplnekA{\leaders\hrule height 0.8pt\hfill}
\def\vyplnekB{\leaders\hrule height 6 pt depth -5.2pt\hfill}
\def\nadpis#1\par{\medbreak \hbox to \hsize{{\color{mygreen}\vyplnekA\ {\textsc{#1}}\vyplnekB}}\par\medbreak}


%\def\vec#1{\boldsymbol{#1}}
\def\norm#1{\left\Vert#1\right\Vert}
\def\x{\norm{x}}
\def\w{\norm{\vec{w}}}
\def\a{{\alpha}}
\def\aa{{\alpha-1}}
\def\at{{a\leq\x\leq t}}
\def\o{\omega_n}
\def\O{\Omega}
\def\c{\cdot}
\def\const{\hbox{const}}
\def\eps{\varepsilon}
\let\epsilon\varepsilon
\interdisplaylinepenalty 50
\setcounter{tocdepth}{1}

\raggedbottom
\let\rmdefault\sfdefault

\usepackage{graphicx}
\usepackage{multicol}

\def\ss#1#2{\left\langle#1,#2\right\rangle}


\makeatletter
\renewcommand\maketitle
{%
    \thispagestyle{empty}%
    \null\bigskip\bigskip
    \ifeqforpaper\vspace*{2\baselineskip}%
    \else
        \vbox to\titleauthorproportion\textheight\bgroup%
    \fi
    \noindent\makebox[\linewidth]{\parbox{\linewidth}%
    {\bfseries\color{\webuniversity@color}\ifeqforpaper\large\fi
    \centering\webuniversity}}\par\ifeqforpaper\else\minimumskip\fi
    \vspace{\stretch{1}}%
    \noindent\makebox[\linewidth]{%
        \parbox{\hproportionwebtitle\linewidth}%
        {\bfseries\color{\webtitle@color}\ifeqforpaper\Large\else
        \large\fi\centering\webtitle}}\par\ifeqforpaper
        \vspace{2\baselineskip}\else\minimumskip\vspace{\stretch{1}}\fi
    \noindent\makebox[\linewidth]{%
        \parbox{\hproportionwebauthor\linewidth}%
        {\bfseries\color{\webauthor@color}\ifeqforpaper
        \large\fi\centering\webauthor}}
    \ifeqforpaper\else
    \egroup % end of \vbox for title and author
    \fi\bigskip
    \optionalpagematter
    \par\vspace{\stretch{1}}
    \ifx\web@directory@option y\webdirectory\fi
    \par\ifeqforpaper\else\minimumskip\fi\vspace{\stretch{1}}
    \vfill\noindent\begingroup
    \trailerFontSize\titlepageTrailer\par\endgroup
    \newpage
}

\makeatother

\pagestyle{empty}%

%\usepackage[inactive]{fancytooltips}
\begin{document}

\def\TooltipRefmark{\hbox{\ \ }}
\def\TooltipExtratext{\hbox{\ \ }}

 \title{Conjugacy criteria for half-linear ODE \\in theory of PDE\\ with
     generalized $p$-Laplacian\\ and mixed powers\\[15pt]}

\author{Robert Ma\v r\'\i k\\[6mm]Dpt. of Mathematics\\ Mendel University\\Brno, CZ 
 }

\date{}

\maketitle

\begin{equation}
  \begin{aligned}
    \div\left(A(x)\norm{\nabla y}^{p-2}\nabla y\right)&+ \ss{\vec
      b(x)}{\norm{\nabla y}^{p-2}\nabla y}\\&+
    c(x)|y|^{p-2}y+\sum_{i=1}^m c_i(x)|y|^{p_i-2}y=e(x),
  \end{aligned}
\tag{E}
\end{equation}
\begin{itemize}
\item $x=(x_1,\ldots,x_n)_{i=1}^n\in\R^n$, $p>1$, $p_i>1$,
\item $A(x)$ is elliptic $n\times n$ matrix with differentiable
  components, $c(x)$ and $c_i(x)$ are H\"older continuous functions,
  $\vec b(x)=\bigl(b_1(x),\ldots,b_n(x)\bigr)$ is continuous
  $n$-vector function,
\item $\nabla=\left({\partial \over\partial
    x_1},\ldots,{\partial \over\partial
    x_n}\right)_{i=1}^n$ and  $\div={\partial \over\partial
    x_1}+\cdots+{\partial \over\partial
    x_n}$ is are the usual nabla and divergence operators,
\item $q$ is a conjugate
  number to the number $p$, i.e., $q=\frac p{p-1}$,
\item $\ss{\cdot}{\cdot}$ is the usual scalar product in $\R^n$,
  $\Vert{\cdot}\Vert$ is the usual norm in $\R^n$, $\Vert A\Vert
  =\sup\left\{\Vert Ax\Vert: x\in \R^n \text{ with } \Vert x\Vert
    =1\right\}=\lambdamax$ is the spectral norm
\item \textbf{solution} of \eqref{eq:E} in $\Omega\subseteq \R^n$ is a
differentiable function $u(x)$ such that $A(x)\Vert\nabla
u(x)\Vert^{p-2}\nabla u(x)$ is also differentiable and $u$ satisfies
\eqref{eq:E} in $\Omega$
\item $ S(a)=\{x\in\R^n: \Vert x\Vert =a\}$, \\$
  \Omega(a)=\{x\in\R^n:a\leq \Vert x\Vert \}$, \\$
  \Omega(a,b)=\{x\in\R^n:a\leq\Vert x\Vert \leq b\}$
\end{itemize}

\newpage

% \begin{equation}
%   {\shadowbox{$\div\Bigl(A(x)\Vert\nabla u\Vert^{p-2}\nabla u\Bigr) + \ss{\vec b(x)}{\Vert\nabla u\Vert^{p-2}\nabla u}+c(x)|u|^{p-2}u=0$}} \tag{E}
% \end{equation}

\nadpis {Concept of oscillation for ODE} 

\begin{equation}
  u''+c(x)u=0 \label{eq1} 
\end{equation}

\begin{itemize}
\item Equation \eqref{eq1} is oscillatory if each solution has
  infinitely many zeros in $[x_0,\infty)$.
\item Equation \eqref{eq1} is oscillatory if each solution has a zero $[a,\infty)$
  for each $a$.
\item Equation \eqref{eq1} is oscillatory if each solution has
  conjugate points on the interval $[a,\infty)$ for each $a$.
\item All definition are equivalent (no accumulation of zeros and
  Sturm separation theorem).
\item Equation is oscillatory if $c(x)$ is large enough. Many
  oscillation criteria are expressed in terms of the integral
  $\int^\infty c(x)\dx$ (Hille and Nehari type)
\item There are oscillation criteria which can detect oscillation even
  if $\int^\infty c(x)\dx$ is extremly small. These criteria are in
  fact series of conjugacy criteria.
\end{itemize}

\newpage
\nadpis Equation with mixed powers

\begin{equation}
  \label{eq:Sun}
  (p(t)u')'+c(t)u+\sum_{i=1}^m c_i(t)|u|^{\alpha _i}\sgn u=e(t)
\end{equation}
where $\alpha_1>\cdots >\alpha_m>1>\alpha_{m+1}>\cdots>\alpha_n>0$.
\begin{Theorem}[Sun,Wong (2007)]
\label{theorem:sun_wong}
  If for any $T\geq 0$ there exists $a_1$, $b_1$, $a_2$, $b_2$ such
  that $T\leq a_1<b_1\leq a_2<b_2$ and
  \begin{equation*}
    \begin{cases}
      c_i(t)\geq 0& t\in[a_1,b_1]\cup[a_2,b_2],\ i=1,2,\dots,n\\
      e(x)\leq 0& t\in[a_1,b_1]\\
      e(x)\geq 0& t\in[a_2,b_2]
    \end{cases}
  \end{equation*}
  and there exists a continuously differentiable function $u(t)$
  satisfying $u(a_i)=u(b_i)=0$, $u(t)\neq 0$ on $(a_i,b_i)$ and
  \begin{equation}\label{eq:SW}
    \int_{a_i}^{b_i}\left\{p(t)u'^2(t)-Q(t)u^2(t)\right\}\dt\leq 0
  \end{equation}
  for $i=1,2$, where
  \begin{equation*}
    Q(t)=k_0|e(t)|^{\eta_0}\prod_{i=1}^m\Bigl(c_i^{\eta_i}(t)\Bigr)+c(t),
  \end{equation*}
  $k_0=\prod_{i=0}^m\eta_i^{-\eta_i}$ and $\eta_i$, $i=0,\dots,n$ are
  positive constants satisfying
%   \begin{equation*}
$    \sum_{i=1}^m\alpha_i\eta_i=1\quad\text{and}\quad \sum_{i=0}^m\eta_i=1$,
%   \end{equation*}
  then all solutions of \eqref{eq:Sun} are oscillatory.
\end{Theorem}



\newpage
\nadpis {Concept of oscillation for linear PDE} 

\begin{equation}
  \Delta u+c(x)u=0 \label{eq2} 
\end{equation}

\begin{itemize}
\item Equation \eqref{eq2} is \textit{oscillatory} if every solution
  has a zero on $\{x\in\R^n: \norm x\geq a\}$ for each $a$.
\item Equation \eqref{eq2} is \textit{nodally oscillatory} if every
  solution has a nodal domain on $\{x\in\R^n: \norm x\geq a\}$ for
  each $a$.
\item Both definition are equivalent (Moss+Piepenbrink).
\end{itemize}


\nadpis {Concept of oscillation for half-linear PDE} 

\begin{equation}
  \div\Bigl(\norm{\nabla u}^{p-2}\nabla u\Bigr)+c(x)|u|^{p-2}u=0 \label{eq3} 
\end{equation}

\begin{itemize}
\item Essentialy the same approach to oscillation as in linear case
\item The equivalence between two oscillations is open problem.
\end{itemize}


% \newpage
% \nadpis Riccati substituion

% If $u$ is a positive solution of the equation
% \begin{equation}\label{eq:linODE}
%   u''+c(x)u=0, 
% \end{equation} then the function
% $w=\frac{u'}{u}$ is a solution of the Riccati type differential equation
% \begin{equation}
% w'+c(x)+|w|^2=0.\label{eq:riceq}
% \end{equation}


% \textbf{Remark:} In fact
% \begin{equation}
% w'+c(x)+|w|^2\leq 0\label{eq:RICineq}
% \end{equation}
% is sufficient in proofs of nonexistence of positive (nonoscillatory)
% solution \eqref{eq:linODE}, since solvability of \eqref{eq:RICineq}
% implies solvability of \eqref{eq:riceq}.


% \nadpis Transforming ODE result (nonexistence of positive solution)
% into PDE

% \null

% \vskip -3\baselineskip

% \null

% % The method used to prove most of oscillation criteria for half-linear PDE
% \begin{enumerate*}
% \item Suppose by contradiction that the PDE possesses positive
%   (eventually positive) solution.
% \item Using transformation
% %   \begin{equation*}
% $    \vec w(x)=
% \frac{\Vert \nabla u(x)\Vert ^{p-2}\nabla u(x)}{|u(x)|^{p-2}u(x)}
% $
% %   \end{equation*}
%   convert positive solutions of
%   \begin{equation*}
%     \div\Bigl(\Vert\nabla u\Vert^{p-2}\nabla u\Bigr)+c(x)|u|^{p-2}u=0
%   \end{equation*}
%   into 
%   \begin{equation}\label{5RIC}
%     \div \vec w+c(x)  +(p-1)\ss{\vec w}{\frac{\nabla u(x)}{u(x)}}=0.
%   \end{equation}
% \item Integrating \eqref{5RIC} over spheres and using standard tools
%   derive a Riccati type inequality of the form \eqref{eq:RICineq} and
%   proceed as in the ODE case.
% \end{enumerate*}

\newpage

\null
\kern-2\baselineskip

\begin{equation}
  \begin{aligned}
    \div\left(A(x)\norm{\nabla y}^{p-2}\nabla y\right)&+ \ss{\vec
      b(x)}{\norm{\nabla y}^{p-2}\nabla y}\\&+
    c(x)|y|^{p-2}y+\sum_{i=1}^m c_i(x)|y|^{p_i-2}y=e(x),
  \end{aligned}
\tag{E}
\end{equation}

\nadpis Detection of oscillation from ODE

% Oscillation of partial differential equation can be detected from
% oscillation of ordinary differential equation.
\begin{Theorem}[O. Do\v sl\'y (2001)] \label{rad}
%   Let
%   \begin{align*}
% %    \~a(r)={1\over \omega_nr^{n-1}}\int_{S(r)}a(x)\dS\\
%     \hat c(r)={1\over \omega_nr^{n-1}}\int_{S(r)}c(x)\dS.
%   \end{align*}
Equation
\begin{equation}
\div(\Vert\nabla u\Vert^{p-2}\nabla u)+c(x)|u|^{p-2}u=0\label{eq:E-non-damp}
\end{equation}
is oscillatory, if the ordinary differential equation
 \begin{equation}
   \label{hl}
   \Bigl( r^{n-1}|u'|^{p-2}u'\Bigr)'+r^{n-1}\left(\frac{1}{\omega_n r^{n-1}}\int_{S(r)}\, c(x) \,\dx\right)|u|^{p-2}u=0
 \end{equation}
is oscillatory.
The number $\omega_n$ is the surface area of the unit sphere in $\R^n$.
\end{Theorem}

J. Jaro\v s, T. Kusano and N. Yoshida proved independently similar
result (for $A(x)=a(\Vert x\Vert )I$, $a(\cdot)$ differentiable).

\nadpis {Our aim}

\begin{itemize*}
\item Extend method used in Theorem \ref{theorem:sun_wong} to
  \eqref{eq:E}. Derive a general result, like Theorem B.
\item Derive a result which does depend on more general expression,
  than the mean value of $c(x)$ over spheres centered in the origin.
%   Is it possible to detect oscillation in such an extreme case as
%   $\int_{S(||x||)}\modra{c(x)}\dS=0$?
\item Remove restrictions used by previous authors (for example Xu (2009)
  excluded the possibility $p_i>p$ for every $i$).
  % S(r)}\cervena{\lambdamax(x)}}$ plays a crucial role in the linear
  % case and $\boxed{\rho(r)\geq \max_{x\in S(r)}\cervena{\frac{\Vert
  %       {A(x)}\Vert ^p_F}{\lambdamin^{p-1}(x)}}}$ plays similar role
  % if $p>1$.  This phenomenon can be observed also in other
  % oscillation criteria than Theorems B and C. We know that
  % $\rho(r)\geq \lambda(r)$. Why such a discrepancy appears?
\end{itemize*}

\newpage
\begin{equation}
  \begin{aligned}
    \div\left(A(x)\norm{\nabla y}^{p-2}\nabla y\right)&+ \ss{\vec
      b(x)}{\norm{\nabla y}^{p-2}\nabla y}\\&+
    \modra{c(x)|y|^{p-2}y}+\cervena{\sum_{i=1}^m c_i(x)|y|^{p_i-2}y}=\cervena{e(x)},
  \end{aligned}
\tag{E}
\label{eq:E}
\end{equation}

\nadpis Modus operandi

\begin{itemize}
\item Get rid of terms $\sum_{i=1}^m c_i(x)|y|^{p_i-2}y$ and $e(x)$
  (join with $c(x)|y|^{p-2}y$) and convert the problem into
  \begin{equation*}
        \div\left(A(x)\norm{\nabla y}^{p-2}\nabla y\right)+ \ss{\vec
      b(x)}{\norm{\nabla y}^{p-2}\nabla y}+\modra{C(x)|y|^{p-2}y}=0.
  \end{equation*}
\item Derive Riccati type inequality in $n$ variables.
\item Derive Riccati type inequality in $1$ variable.
\item Use this inequality as a tool which transforms results from ODE
  to PDE.
\end{itemize}


\newpage

Using generalized AG inequality $\sum \alpha _i\geq
\prod\left(\frac{\alpha_i}{\eta_i}\right)^{\eta_i}$, if $\alpha_i\geq
0$, $\eta_i>0$ and $\sum \eta_i=1$ we eliminate the right-hand side and terms with mixed powers.



\begin{lemma}\label{lemma:est1}
  Let either $y>0$ and $e(x)\leq 0$ or $y<0$ and $e(x)\geq 0$. Let
  $\eta_i>0$ be numbers satisfying $\sum_{i=0}^m{\eta_i}=1$ and
  $\eta_0+\sum_{i=1}^m p_i\eta_i=p$ and let $c_i(x)\geq 0$ for every
  $i$. Then
  \begin{equation*}%\label{eq:est1}
    \frac{1}{|y|^{p-2}y}\left(-e(x)+\sum_{i=1}^m c_i(x)|y|^{p_i-2} y\right)\geq C_1(x),
  \end{equation*}
  where 
  \begin{equation}
    \label{eq:C1}
     C_1(x):=\left|\frac{e(x)}{\eta_0}\right|^{\eta_0} 
    \prod_{i=1}^m\left(\frac{c_i(x)}{\eta_i}\right)^{\eta_i}.
  \end{equation}
\end{lemma}



%\begin{remark}
\textbf{Remark:} The numbers $\eta_i$ from Lemma \ref{lemma:est1} exist, if $p_i>p$ for some $i$.
%\end{remark}

% The following lemma is a modification of Lemma \ref{lemma:est1} in
% the case $e(x)\equiv 0$.

\begin{lemma}\label{lemma:est10}
  Suppose $c_i(x)\geq 0$. Let $\eta_i>0$ be numbers satisfying
  $\sum_{i=1}^m{\eta_i}=1$ and $\sum_{i=1}^m p_i\eta_i=p$. Then
  \begin{equation*}%\label{eq:est10}
        \frac{1}{|y|^{p-2}y}\sum_{i=1}^m c_i(x)|y|^{p_i-2}y\geq C_2(x),
  \end{equation*}
  where
  \begin{equation}
    \label{eq:C2}
        C_2(x):=\prod_{i=1}^m\left(\frac{c_i(x)}{\eta_i}\right)^{\eta_i}
  \end{equation}
\end{lemma}

% \begin{remark}
\textbf{Remark:} The numbers $\eta_i$ from Lemma \ref{lemma:est10}
exist iff $p_i>p$ for some $i$ and $p_j<p$ for some $j$.
% \end{remark}


\newpage
% \begin{lemma}\label{lemma:ineq_cal}
%   The following inequalities hold for $a\geq 0$ and $x>0$.
%   \begin{enumerate}
%   \item If $\alpha<\beta$ and $b>0$, then $b-ax^\alpha\geq -x^\beta \left(\frac{a(\beta-\alpha)}{b\beta}\right)^{\frac\beta\alpha} \frac{b\alpha}{\beta-\alpha}$
%   \label{pa}
%   \item If $\alpha>\beta$ and $b\geq0$, then $ax^\alpha+b\geq x^\beta \left(\frac{a(\alpha-\beta)}{b\beta}\right)^{\frac\beta\alpha} \frac{b\alpha}{\alpha-\beta}$
%   \label{pb}
%   \end{enumerate}  
% \end{lemma}

% Another possibility how to remove the right hand side and terms with
% mixed powers is available if we rewrite
% \begin{equation*}
%   \frac{1}{|y|^{p-2}y}\left(-e(x)+\sum_{i=1}^m c_i(x)|y|^{p_i-2}y\right)
% \end{equation*}
% into the form 
% \begin{equation*}
%   \sum_{i=1}^m \left(c_i(x)|y|^{p_i-p}-\frac{\epsilon_i e(x)}{|y|^{p-2}y} \right), \quad \epsilon_i>0, \quad \sum_{i=1}^m\epsilon_i=1
% \end{equation*}
% study the family of min/max problems
% for terms in this sum.

% \bigskip

% \begin{lemma}\label{lemma:estimate2}
%   Let $e(x)<0$ and $y>0$. Then
% \begin{equation*}%\label{eq:estimate2}
% \sum_{i=1}^m c_i(x)|y|^{p_i-p}-\frac{e(x)}{|y|^{p-2}y}
% \geq C_3(x),
% \end{equation*}
% where 
% \begin{multline}
%   \label{eq:C3}
%   C_3(x):=\sum_{i\in I_1}
% \left(\left[\frac{[c_i(x)]_+(p_i-p)}{\epsilon_i|e(x)|(p-1)}\right]^{(p-1)/(p_i-1)}\frac{\epsilon_i|e(x)|(p_i-1)}{p_i-p}\right)\\
% -   \sum_{i\in I_2}\left(\left[\frac{[-c_i(x)]_+(p-p_i)}{\epsilon_i|e(x)|(p-1)}\right]^{(p-1)/(p_i-1)}\frac{\epsilon_i|e(x)|(p_i-1)}{p-p_i}\right),
% \end{multline}
% $I_1=\{i\in[1,m]\cap \N:p_i>p\}$ and $I_2=\{i\in[1,m]\cap \N:p_i<p\}$,
% $\epsilon_i>0$, $\sum_{i=1}^m\epsilon_i=1$. Moreover, if
% $I_2=\{\}$, then the inequality $e(x)<0$ can be relaxed to
% $e(x)\leq 0$.
% \end{lemma}



% \newpage
\begin{lemma}\label{lemma:cC}
  Let $y$ be a solution of \eqref{eq:E} which does not have zero on
  $\Omega$. Suppose that there exists a function 
  $C(x)$ such that 
  \begin{equation*}
    C(x)\leq c(x)+\sum_{i=1}^m c_i(x)|y|^{p_i-p}-\frac{e(x)}{|y|^{p-2}y}
%    \label{ineq:C}
  \end{equation*}
  Denote $\vec w(x)=A(x)\frac{\norm{\nabla y}^{p-2}\nabla
    y}{|y|^{p-2}y}$. The function $\vec w(x)$ is well defined on
  $\Omega$ and satisfies the inequality
  \begin{equation}
    \label{eq:RIC}
    \div \vec w+(p-1)\Lambda(x) \norm{\vec w}^q+\ss{\vec w}{A^{-1}(x)\vec b(x)}+C(x)\leq 0
  \end{equation}
  where
  \begin{equation}\label{eq:Lambda}
    \Lambda(x)=
    \begin{cases}
      \lambda_{{\max}}^{1-q}(x)& % \text{ for }
      1<p\leq 2,\\
      \lambda_{{\min}}\lambda_{\max}^{-q}(x)&  % \text{ for } 
      p>2.
      \end{cases}
    \end{equation}
  \end{lemma}

\begin{lemma}\label{lemma:alpha}
  Let \eqref{eq:RIC} hold. Let $l>1$, $l^*=\frac{l}{l-1}$ be two
  mutually conjugate numbers and $\alpha \in C^1(\Omega,\R^+)$ be
  a smooth function positive on $\Omega$.  Then
    \begin{multline*}
 %   \label{eq:RIC2}
    \div (\alpha(x)\vec w)+ (p-1)\frac {\Lambda(x)\alpha^{1-q}(x)}{l^*}
    \norm{\alpha(x)\vec w}^q\\
    -\frac{l^{p-1}\alpha(x)}{ p^p \Lambda^{p-1}(x)}\norm{A^{-1}(x)\vec b(x)-\frac{\nabla \alpha(x)}{\alpha(x)}}^p +\alpha(x)C(x)\leq 0
  \end{multline*}
  holds on $\Omega$.  If $\norm{A^{-1}\vec b-\frac{\nabla
      \alpha}\alpha}\equiv 0$ on $\Omega$, then this inequality holds
  with $l^*=1$.
\end{lemma}

\newpage
\begin{theorem}\label{lemma:radialODE}
  Let the $n$-vector function $\vec w$ satisfy inequality
  \begin{equation*}
    \div \vec w+C_0(x)+(p-1)\Lambda_0(x)\norm{\vec w}^q\leq 0
  \end{equation*}
  on $\Omega(a,b)$.  Denote $\tilde C(r)=\int_{S(r)}C_0(x)\dS$ and
  $\tilde R(r)=\int_{S(r)}\Lambda_0^{1-p}\dS$.  Then
  the half-linear ordinary differential equation
  \begin{equation*}%\label{eq:radialODE}
    \left(\tilde R(r) |u'|^{p-2}u\right)'+\tilde C(r) |u|^{p-2}u=0, 
    \qquad {}'=\frac{\mathrm{d}}{\dr}
  \end{equation*}
  is disconjugate on $[a,b]$ and it possesses solution which has no
  zero on $[a,b]$.
\end{theorem}

\begin{theorem}\label{th1}
  Let $l>1$. Let  $l^*={1}$ if $\norm{\vec b}\equiv 0$ and
  $l^*=\frac{l}{l-1}$ otherwise. Further, let $c_i(x)\geq 0$ for every
  $i$. Denote
  \begin{equation*}%\label{eq:tildeR}
    \tilde R(r)=(l^*)^{p-1}\int_{S(r)}\Lambda^{1-p}(x)\dS 
  \end{equation*}
  and 
  \begin{equation*}
    \tilde C(r)=\int_{S(r)}c(x)+C_1(x)-\frac{l^{p-1}}{ p^p \Lambda^{p-1}(x)}\norm{A^{-1}(x)\vec b(x)}^p\dS,
  \end{equation*}
  where $\Lambda(x)$ is defined by \eqref{eq:Lambda} and $C_1(x)$ is
  defined by \eqref{eq:C1}.

  Suppose that the equation
  \begin{equation*}%\label{eq:th1}
    \left(\tilde R(r)|u'|^{p-2}u'\right)'+\tilde C(r) |u|^{p-2}u=0
  \end{equation*}
  has conjugate points on $[a,b]$.  
  
  If $e(x)\leq 0$ on $\Omega(a,b)$, then equation \eqref{eq:E} has no
  positive solution on $\Omega(a,b)$.

  If $e(x)\geq 0$ on $\Omega(a,b)$, then equation \eqref{eq:E} has no
  negative solution on $\Omega(a,b)$.
\end{theorem}

\begin{theorem}[non-radial variant of Theorem \ref{th1}]\label{th1a}
  Let $l>1$ and let $\Omega\subset\Omega(a,b)$ be an open domain with
  piecewise smooth boundary such that $\meas(\Omega \cap S(r))\neq 0$
  for every $r\in[a,b]$. Let $c_i(x)\geq 0$ on $\Omega$ for every
  $i$ and let $\alpha(x)$ be a function which is
  positive and continuously differentiable on $\Omega$ and vanishes on
  the boundary and outside $\Omega$.  Let $l^*=1$ if $\norm{A^{-1}\vec
    b-\frac{\nabla \alpha}{\alpha}}\equiv 0$ on $\Omega$ and
  $l^*=\frac{l}{l-1}$ otherwise. In the former case suppose also that
  the integral
  \begin{equation*}
    \int_{S(r)}\frac{\alpha(x)}{ \Lambda^{p-1}(x)}\norm{A^{-1}(x)\vec b(x)-\frac{\nabla\alpha(x)}{\alpha(x)}}^p\dS
  \end{equation*}
  which may have singularity on $\partial \Omega$ if
  $\Omega\neq\Omega(a,b)$ is convergent for every $r\in[a,b]$. Denote
  \begin{equation*}
    \tilde R(r)=(l^*)^{p-1}\int_{S(r)}\alpha(x)\Lambda^{1-p}(x)\dS 
  \end{equation*}
  and
  \begin{equation*}
    \tilde C(r)=\int_{S(r)}{\cervena{\alpha(x)}}\left(c(x)+C_1(x)-\frac{l^{p-1}}{ p^p \Lambda^{p-1}(x)}\norm{A^{-1}(x)\vec b(x)-\frac{\nabla\alpha(x)}{\alpha(x)}}^p\right)\dS,
  \end{equation*}
  where $\Lambda(x)$ is defined by \eqref{eq:Lambda} and $C_1(x)$ is
  defined by \eqref{eq:C1} and suppose that equation
  \begin{equation*}
    \left(\tilde R(r)|u'|^{p-2}u'\right)'+\tilde C(r) |u|^{p-2}u=0
  \end{equation*}
  has conjugate points on $[a,b]$.
  
  If $e(x)\leq 0$ on $\Omega(a,b)$, then equation \eqref{eq:E} has no
  positive solution on $\Omega(a,b)$.

  If $e(x)\geq 0$ on $\Omega(a,b)$, then equation \eqref{eq:E} has no
  negative solution on $\Omega(a,b)$.  
\end{theorem}

\newpage

\begin{theorem}\label{th2}
  Let $l$, $\Omega$, $\alpha(x)$, $\Lambda(x)$ and $\tilde R(r)$ be
  defined as in Theorem \ref{th1a} and let $c_i(x)\geq 0$ and
  \cervena{$e(x)\equiv 0$} on $\Omega(a,b)$. Denote
  \begin{equation*}
    \tilde C(r)=\int_{S(r)}\alpha(x)\left(c(x)+C_2(x)-\frac{l^{p-1}}{ p^p \Lambda^{p-1}(x)}\norm{A^{-1}(x)\vec b(x)-\frac{\nabla \alpha(x)}{\alpha(x)}}^p\right)\dS,
  \end{equation*}
  where $C_2(x)$ is defined by \eqref{eq:C2}.
  If the equation %\eqref{eq:th1}
   \begin{equation*}
     \left(\tilde R(r)|u'|^{p-2}u'\right)'+\tilde C(r) |u|^{p-2}u=0
   \end{equation*}
  has conjugate points on $[a,b]$, then every solution of equation
  \eqref{eq:E} has zero on $\Omega(a,b)$.
\end{theorem}

\bigskip\bigskip\bigskip
{\rightskip 2cm
\leftskip 2cm

Similar theorems can be derived also for estimates of terms
with mixed powers based on different methods than AG inequality % (for example
% \eqref{eq:C3})
(see R. M., Nonlinear Analysis TMA 73 (2010)).

}

\end{document}