Research Article Exponential Attractors for Parabolic Equations with Dynamic Boundary Conditions Zhao-hui Fan

advertisement
Hindawi Publishing Corporation
Journal of Applied Mathematics
Volume 2013, Article ID 389863, 6 pages
http://dx.doi.org/10.1155/2013/389863
Research Article
Exponential Attractors for Parabolic Equations with Dynamic
Boundary Conditions
Zhao-hui Fan
Department of Mathematics, Zhengzhou University, Zhengzhou, Henan 450001, China
Correspondence should be addressed to Zhao-hui Fan; fanzhh02@163.com
Received 24 December 2012; Accepted 2 February 2013
Academic Editor: Debasish Roy
Copyright © 2013 Zhao-hui Fan. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We study exponential attractors for semilinear parabolic equations with dynamic boundary conditions in bounded domains. First,
we give the existence of the exponential attractor in 𝐿2 (Ω) × πΏ2 (Γ) by proving that the corresponding semigroup satisfies the
enhanced flattering property. Second, we apply asymptotic a priori estimate and obtain the exponential attractor in 𝐿𝑝 (Ω) × πΏπ‘ž (Γ).
Finally, we show the exponential attractor in (𝐻1 (Ω) ∩ 𝐿𝑝 (Ω)) × πΏπ‘ž (Γ).
1. Introduction
Parabolic equations with dynamical boundary conditions
have strong backgrounds in mathematical physics. They
arise in the heat transfer theory in a solid in contact with
moving fluid, thermoelasticity, diffusion phenomena, heat
transfer in two medium, problems in fluid dynamic, and so
forth. At present, there are many monographs in the whole
world (see [1–13]). Several approaches have been used for
these equations, like the theory of semigroup, with Bessel
potential and Besov space, and the variational setting. In
particular, we are devoted to the long-time behavior of the
solutions. For instance, In [1], the authors showed existence of
pullback attractors. In [8, 9], the authors gave well posedness
and global attractors in 𝐿𝑝 (Ω) × πΏπ‘ (Γ). In [12, 13]; the
authors obtained uniform attractors and some asymptotic
regularity of global attractors in (𝐻1 (Ω) ∩ 𝐿𝑝 (Ω)) × πΏπ‘ž (Γ).
An exponential attractor, in contrast to a global attractor
(or a uniform attractor), enjoys a uniform exponential rate
of convergence of its solution. Because of this, exponential attractors possess more practical property. But to our
knowledge, it does not seem to be in the literature any study
of the existence of exponential attractors for this kind of
equations.
This paper is concerned with existence of exponential
attractors for the following reaction-diffusion equation with
dynamic boundary condition
𝑒𝑑 − Δ𝑒 + 𝑓 (𝑒) = β„Ž (π‘₯)
in Ω,
πœ•π‘’
+ 𝑔 (𝑒) = 𝜌 (π‘₯)
πœ•]
on Γ,
𝑒𝑑 +
𝑒 (0, π‘₯) = 𝑒0 (π‘₯) ,
(1)
in Ω,
where Ω ⊂ R𝑁, 𝑁 ≥ 1, is a bounded domain with a smooth
boundary Γ. Here ] is the outer unit normal on Γ.
In [14], the authors established some necessary and sufficient conditions for the existence of exponential attractors
for continuous and norm-to-weak continuous semigroup
and provided a new method for proving the existence
of exponential attractors by combining with the flattering
property. Motivated by some ideas in [14–16], we combine
asymptotic a prior estimate with the enhanced flattening
property and show sufficient and necessary existence of
exponential attractors in uniformly convex Banach spaces.
As an application, we prove the existence of exponential
attractors for the reaction-diffusion equation with dynamic
boundary condition.
This paper is organized as follows. In Section 2, we
recall some basic results and then give our theorems,
that is, Theorems 5, 6, and 9 and the solution semigroup
corresponding to (1). In Section 3, we obtain the exponential
2
Journal of Applied Mathematics
attractor in 𝐿2 (Ω) × πΏ2 (Γ) for weak solutions, then combine
asymptotical a prior estimate and show the exponential
attractor in 𝐿𝑝 (Ω) × πΏπ‘ž (Γ). Finally, we derive the existence
of the exponential attractor in the space (𝐻1 (Ω) ∩ 𝐿𝑝 (Ω)) ×
πΏπ‘ž (Γ).
In the following, the constants 𝐢𝑖 , 𝑖 = 0, 1, 2, . . . will
always denote generic constants different in various occurrences. The symbol 𝐢(𝑠) will denote a positive constant
dependent of 𝑠. We will write Ω𝑇 for Ω × (0, 𝑇), Γ𝑇 for
Γ × (0, 𝑇), (⋅, ⋅) for the inner product in 𝐿2 (Ω) and ⟨⋅, ⋅⟩ for
the inner product in 𝐿2 (Γ). For convenience, we denote the
norm of 𝑒 by |𝑒|Ω in the space 𝐿2 (Ω) and |𝑒|Γ in the space
𝐿2 (Γ).
2. Preliminary
2.1. The Basic Results and Theorems. Let 𝑋 be a complete
metric space and a one-parameter family of mappings 𝑆(𝑑):
let 𝑋 → 𝑋(𝑑 ≥ 0) be a semigroup. Here we omit the definitions of continuous or norm-to-weak semigroups, dynamical
systems, global attractors, and exponential attractors (see [15–
20]).
Definition 1. Let 𝑋 be a metric space and 𝐡 be a bounded
subset of 𝑋. The Kuratowski measure of noncompactness
𝛼(𝐡) of 𝐡 is defined as
𝛼 (𝐡) = inf {𝛿 > 0 | 𝐡 admits a finite cover by sets
of diameter ≤ 𝛿} .
projector, π‘š is the dimension of 𝑋1 , || ⋅ || denotes
the norm in 𝑋, and π‘˜(𝑠) is a real-valued function
satisfying lim𝑠 → ∞ π‘˜(𝑠) = 0.
Inspired by [14, 21], we easily obtain the following.
Theorem 5. Let 𝑋 be a uniformly convex Banach space
and {𝑆(𝑑)}𝑑≥0 be a continuous or norm-to-weak continuous
semigroup in 𝑋. Then the following conditions are equivalent.
(1) The measure of noncompactness is exponentially decaying for dynamical system (𝑆(𝑑), 𝑋), that is, there exist
π‘˜, 𝑙 > 0 such that 𝛼(⋃𝑠>𝑑 𝑆(𝑠)𝐡) ≤ π‘˜π‘’−𝑙𝑑 .
(2) 𝑆(𝑑) satisfies the enhanced flattening property.
Proof. (1 ⇒ 2) On account of condition (1), for any bounded
subset 𝐡 of 𝑋 and for any πœ€ > 0, there exist 𝑑 = 𝑑(𝐡) such that
𝛼 (⋃𝑆 (𝑠) 𝐡) ≤ π‘˜π‘’−𝑙𝑑 .
(3)
𝑠>𝑑
Namely, there exist a finite number of subset 𝐴 1 , 𝐴 2 , . . . , 𝐴 𝑛
with diameter less than π‘˜π‘’−𝑙𝑑 , such that
⋃𝑆 (𝑠) 𝐡 ⊂ ∪𝑛𝑖=1 𝐴 𝑖 .
(4)
⋃𝑆 (𝑠) 𝐡 ⊂ ∪𝑛𝑖=1 𝑁 (π‘₯𝑖 , π‘˜π‘’−𝑙𝑑 ) .
(5)
𝑠>𝑑
Let π‘₯𝑖 ∈ 𝐴 𝑖 , then
(2)
Theorem 2 (see [14]). Assume that 𝐡 is a bounded absorbing
set for discrete dynamical system 𝑆(𝑛) in 𝑋; then the following
are equivalent.
(1) The measure of noncompactness is exponentially decaying for dynamical system (𝑆(𝑛), 𝑋), that is, there exist
π‘˜, 𝑙 > 0 such that 𝛼(β‹ƒπ‘š>𝑛 𝑆(π‘š)𝐡) ≤ π‘˜π‘’−𝑙𝑛 .
(2) For 𝑆(𝑛), there exist exponential attractors.
Theorem 3 (see [14]). Assume that B is a bounded absorbing
set for 𝑆(𝑑) in 𝑋; then the following are equivalent.
(1) The measure of noncompactness is exponentially decaying for dynamical system (𝑆(𝑑), 𝑋), that is, there exist
π‘˜, 𝑙 > 0 such that 𝛼(⋃𝑠>𝑑 𝑆(𝑠)𝐡) ≤ π‘˜π‘’−𝑙𝑑 .
(2) For 𝑆(𝑑), there exist exponential attractors.
Definition 4 (see [14] (Enhanced Flattening Property)). Let 𝑋
be a uniformly convex Banach space; for any bounded set 𝐡
of 𝑋, there exist π‘˜, 𝑙 > 0 and 𝑇 > 0, and a finite dimension
subspace 𝑋1 of 𝑋, such that
(1) π‘ƒπ‘š (⋃𝑠≥𝑑 𝑆(𝑠)𝐡) is bounded and
(2) ||(𝐼 − π‘ƒπ‘š )(⋃𝑠≥𝑑 𝑆(𝑠)π‘₯)|| ≤ π‘˜π‘’−𝑙𝑑 + π‘˜(π‘š), for all π‘₯ ∈ 𝐡,
for all 𝑑 ≥ 𝑇.Here π‘ƒπ‘š : 𝑋 → 𝑋1 is a bounded
𝑠>𝑑
Let 𝑋1 = span{π‘₯1 , π‘₯2 , . . . , π‘₯𝑛 }, since 𝑋 is uniformly convex,
there exist a projection 𝑃 : 𝑋 → 𝑋1 , such that for any π‘₯ ∈ 𝑋,
||π‘₯ − 𝑃π‘₯|| = dist(π‘₯, 𝑋1 ). Hence,
β€–(𝐼 − 𝑃) 𝑆 (𝑑) π‘₯β€– ≤ π‘˜π‘’−𝑙𝑑 < π‘˜π‘’−𝑙𝑑 + π‘˜ (π‘š) .
(6)
(2 ⇒ 1) have been shown in Theorem 4.3 of [14], so we
omitted it here.
By Theorem 5, we can deduce the following.
Theorem 6. Let 𝑋 be a uniformly convex Banach space
and {𝑆(𝑑)}𝑑≥0 be a continuous or norm-to-weak continuous
semigroup in X. Then, for dynamical system (𝑆(𝑑), 𝑋), there
exist exponential attractors in 𝑋 if and only if
(1) there is a bounded absorbing set 𝐡 ⊂ 𝑋, and
(2) S(t) satisfies the enhanced flattening property.
In addition, we use later the following theorem about
global attractors.
Theorem 7 (see [15, Corollary 5.7]). Let {𝑆(𝑑)}𝑑≥0 be a semigroup on 𝐿𝑝 (Ω), (𝑝 ≥ 1), be a continuous or weak continuous
semigroup on πΏπ‘ž (Ω) for some π‘ž ≤ 𝑝, and have a global attractor
in πΏπ‘ž (Ω). Then, {𝑆(𝑑)}𝑑≥0 has a global attractor in 𝐿𝑝 (Ω) if and
only if
Journal of Applied Mathematics
3
(1) {𝑆(𝑑)}𝑑≥0 has a bounded absorbing set 𝐡0 in 𝐿𝑝 (Ω), and
𝑝
(2) for any πœ€ > 0 and any bounded subset 𝐡 ⊂ 𝐿 (Ω), there
exist positive constants 𝑀 = 𝑀(, 𝐡) and 𝑇 = 𝑇(, 𝐡),
such that
󡄨󡄨
󡄨𝑝
󡄨󡄨𝑆 (𝑑) 𝑒0 󡄨󡄨󡄨 < πœ€
Ω(|𝑆(𝑑)𝑒 |≥𝑀)
∫
for any 𝑒0 ∈ 𝐡, 𝑑 ≥ 𝑇,
(7)
0
From Lemma 8 and Condition (2), it follows that 𝑆(𝑛0 +𝑛1 )(𝐡)
has a finite πœ€/2-net in 𝐿𝑝 (Ω).
By induction, let πœ€/2𝑖−1 be replaced by πœ€/2𝑖 , there 𝑛𝑖 ∈ N
and 𝑀𝑖+1 = 𝐢(πœ€/2𝑖 )𝛾/(π‘ž−𝛾) (𝛾 > π‘ž) such that
(π‘ž−𝑝)/π‘ž
π‘˜π‘’−𝑙(𝑛0 +𝑛1 +⋅⋅⋅+𝑛𝑖 ) ≤ (3𝑀𝑖+1 )
≤ π‘˜π‘’
We give the following lemma concerning the covering of
the set in two different topologies used later in the proof of
Theorem 9.
Lemma 8 (see [15, Lemma 5.3]). For any πœ€ > 0, the bounded
subset B of 𝐿𝑝 (Ω) has a finite πœ€-net in 𝐿𝑝 (Ω) if there exists a
positive constant 𝑀 = 𝑀(πœ€), such that
(1) 𝐡 has a finite (3𝑀)(π‘ž−𝑝)/π‘ž (πœ€/2)𝑝/π‘ž -net in πΏπ‘ž (Ω);
(π‘ž−𝑝)/π‘ž
(3𝑀𝑖+1 )
(1) 𝑆(𝑑) has an exponential attractor in πΏπ‘ž (Ω);
for any 𝑒0 ∈ 𝐡0 ,
(8)
0
then 𝑆(𝑑) has an exponential attractor in 𝐿𝑝 (Ω).
Proof. Take 𝑇 > 0, and let 𝑆𝑛 = 𝑆(𝑛𝑇); obviously 𝑆𝑛 is
a discrete dynamical system. On account of Condition (1),
for 𝑆𝑛 there exists an exponential attractor M in πΏπ‘ž (Ω).
By Theorem 2, we find that there exist π‘˜, 𝑙 > 0 such that
𝛼(β‹ƒπ‘š>𝑛 𝑆(π‘š)𝐡) ≤ π‘˜π‘’−𝑙𝑛 . By the definition of the measure
of noncompactness, for all 𝑛 ∈ 𝑁, there exist finite points
𝑀
π‘₯𝑛𝑖 ∈ 𝑆(𝑛)𝐡 such that 𝑆(𝑛)𝐡 ⊂ ∪𝑖=1𝑛 (π΅πΏπ‘ž (Ω) (π‘₯𝑛𝑖 , π‘˜π‘’−𝑙𝑛 ) ∩ 𝑆(𝑛)𝐡).
Then, there exist 𝑛0 ∈ N and 𝑀1 = πΆπœ€π›Ύ/(π‘ž−𝛾) (𝛾 > π‘ž) such that
(π‘ž−𝑝)/π‘ž
π‘˜π‘’−𝑙𝑛0 ≤ (3𝑀1 )
πœ€ 𝑝/π‘ž
( ) ≤ π‘˜π‘’−𝑙(𝑛0 −1) .
2
(9)
From Lemma 8 and Condition (2), it follows that 𝑆(𝑛0 )(𝐡) has
a finite πœ€-net in 𝐿𝑝 (Ω).
Let πœ€ be replaced by πœ€/2, there 𝑛1 ∈ N and 𝑀2 =
𝐢(πœ€/2)𝛾/(π‘ž−𝛾) (𝛾 > π‘ž) such that
(π‘ž−𝑝)/π‘ž
π‘˜π‘’−𝑙(𝑛0 +𝑛1 ) ≤ (3𝑀2 )
(
πœ€ 𝑝/π‘ž
) ≤ π‘˜π‘’−𝑙(𝑛0 +𝑛1 −1) .
22
(10)
.
(π‘ž−𝑝)/π‘ž
(3𝑀𝑖 )
𝑝/π‘ž
𝑝/π‘ž
(πœ€/2𝑖 )
(𝑀𝑖+1 /𝑀𝑖 )
=
2𝑝/π‘ž
=
((πœ€/2 )
(π‘ž−𝑝)/π‘ž
(π‘ž−𝑝)/π‘ž
𝑖−1 𝛾/(π‘ž−𝛾)
/(πœ€/2
)
(12)
)
2𝑝/π‘ž
1 𝛾(𝑝−π‘ž)/π‘ž(𝛾−π‘ž)+𝑝/π‘ž
=( )
< 1.
2
and
So,
the
ratio
of
(3𝑀𝑖+1 )(π‘ž−𝑝)/π‘ž (πœ€/2𝑖+1 )𝑝/π‘ž
(π‘ž−𝑝)/π‘ž
𝑖 𝑝/π‘ž
(3𝑀𝑖 )
(πœ€/2 )
is a constant for given 𝛾, 𝑝, π‘ž, and
πœ€. In fact,
(π‘ž−𝑝)/π‘ž
(2) for any πœ€ > 0, there exist positive constant 𝑀 = 𝑀(πœ€)
such that
(11)
(πœ€/2𝑖+1 )
𝑖 𝛾/(π‘ž−𝛾)
Theorem 9. Assume that 𝑝 > π‘ž > 0 and Ω ⊂ R𝑛 . Let S(t) be a
continuous or norm-to-weak continuous semigroup on 𝐿𝑝 (Ω)
and πΏπ‘ž (Ω) and 𝐡0 be a positively invariant bounded absorbing
set in 𝐿𝑝 (Ω). If the following conditions hold true:
𝑝/π‘ž
From Lemma 8 and Condition (2), it follows that 𝑆(𝑛0 + 𝑛1 +
⋅ ⋅ ⋅ + 𝑛𝑖 )(𝐡) has a finite πœ€/2𝑖 -net in 𝐿𝑝 (Ω). Note that
(2) ∫Ω(|𝑒|≥𝑀) |𝑒|𝑝 ≤ 2−2𝑝+2 πœ€, for any 𝑒0 ∈ 𝐡0 .
Inspired by [16], we give the subsequent theorem which
describes our new technique to construct an exponential
attractor in a stronger topological space.
πœ€
)
2𝑖+1
−𝑙(𝑛0 +𝑛1 +⋅⋅⋅+𝑛𝑖 −1)
where Ω(|𝑒| ≥ 𝑀) = {π‘₯ ∈ Ω||𝑒(π‘₯)| ≥ 𝑀}.
󡄨󡄨
󡄨𝑝
−2𝑝+2
πœ€,
∫
󡄨󡄨𝑆 (𝑑) 𝑒0 󡄨󡄨󡄨 ≤ 2
Ω(|𝑆(𝑑)𝑒 |≥𝑀)
(
(3𝑀𝑖+1 )
(
πœ€
)
2𝑖+1
𝑝/π‘ž
1 𝑝/π‘ž
= ( ) (3𝐢)(π‘ž−𝑝)/π‘ž πœ€(𝛾(𝑝−π‘ž)/π‘ž(𝛾−π‘ž))+𝑝/π‘ž
2
(13)
1 ((𝛾(𝑝−π‘ž)/π‘ž(𝛾−π‘ž))+𝑝/π‘ž)𝑖
×( )
.
2
Substituting (13) into (11), we deduce easily that
πœ€
(𝑝−π‘ž)/π‘ž π‘ž/𝑝 −π‘žπ‘™(𝑛0 +𝑛1 +⋅⋅⋅+𝑛𝑖 −1)/𝑝
≤ 2(3𝑀𝑖+1 )
π‘˜ 𝑒
2𝑖
1 (𝛾(𝑝−π‘ž)/π‘ž(𝛾−π‘ž))𝑖
= 2(3𝐢)(π‘ž−𝑝)/π‘ž πœ€π›Ύ(𝑝−π‘ž)/π‘ž(𝛾−π‘ž) ( )
2
(14)
× π‘˜π‘ž/𝑝 𝑒−π‘žπ‘™(𝑛0 +𝑛1 +⋅⋅⋅+𝑛𝑖 −1)/𝑝 .
Choose that 2(3𝐢)(π‘ž−𝑝)/π‘ž πœ€π›Ύ(𝑝−π‘ž)/π‘ž(𝛾−π‘ž) π‘˜π‘ž/𝑝 π‘’π‘žπ‘™/𝑝 = π‘˜σΈ€  ≥
2(3𝐢)(π‘ž−𝑝)/π‘ž πœ€π›Ύ(𝑝−π‘ž)/π‘ž(𝛾−π‘ž) (1/2)(𝛾(𝑝−π‘ž)/π‘ž(𝛾−π‘ž))𝑖 π‘˜π‘ž/𝑝 π‘’π‘žπ‘™/𝑝 and 𝑙󸀠 =
π‘žπ‘™/𝑝, we know that 𝑆(𝑛0 + 𝑛1 + ⋅ ⋅ ⋅ + 𝑛𝑖 ) (𝐡) has a finite
σΈ€ 
π‘˜σΈ€  𝑒−𝑙 (𝑛0 +𝑛1 +⋅⋅⋅+𝑛𝑖 ) -net in 𝐿𝑝 (Ω).
Combining (9), (10) (11), and (13), we can choose 𝑛1 =
𝑛2 = ⋅ ⋅ ⋅ = 𝑛𝑖 = [(1/𝑙)(𝛾(𝑝 − π‘ž)/π‘ž(𝛾 − π‘ž) + 𝑝/π‘ž) ln 2] + 1,
where [π‘Ž] is the integer part of π‘Ž. Denoted by 𝑙0 = [(1/𝑙)(𝛾(𝑝−
π‘ž)/π‘ž(𝛾 − π‘ž) + 𝑝/π‘ž) ln 2] + 1, we have 𝑆(𝑛0 + 𝑛1 + ⋅ ⋅ ⋅ +
σΈ€ 
𝑛𝑖 )(𝐡) = 𝑆(𝑛0 + 𝑖𝑙0 ) has a finite π‘˜σΈ€  𝑒−𝑙 (𝑛0 +𝑖𝑙0 ) -net in 𝐿𝑝 (Ω).
By Theorem 2, an exponential attractor exists in 𝐿𝑝 (Ω) for
discrete semigroup 𝑆(𝑖𝑙0 ). Using the same argument as in [18],
4
Journal of Applied Mathematics
it is easy to deduce that an exponential attractor M exists in
𝐿𝑝 (Ω) for discrete semigroup 𝑆(𝑛). Let M0 = ∪0≤𝑠≤𝑇 𝑆(𝑠)M.
As an repetition of the general method developed by Li et
al. [14], we can show that M0 is the exponential attractor for
dynamical system (𝑆(𝑑), 𝑋).
2.2. The Solution Semigroup. We can write Problem (1) as an
evolution for unknown 𝑒(π‘₯, 𝑑) in Ω and V(π‘₯, 𝑑) on Γ
𝑀𝑑 + 𝐴𝑀 + 𝐹 = 0,
(15)
with the compatibility V = 𝛾(𝑒(𝑑)) (𝛾 is trace operator) for
𝑑 > 0 and πœ† > 0, where
𝑒
𝑀 = ( ),
V
−Δ + πœ† 0
),
𝐴=( πœ•
0
πœ•]
(16)
𝑓 (𝑒) − πœ†π‘’ − β„Ž (π‘₯)
).
𝐹=(
𝑔 (V) − 𝜌 (π‘₯)
In the case ٠bounded, the operator 𝐴 has a compact
resolvent and its spectrum, denoted by 𝜎(𝐴) = {πœ‡π‘› }𝑛 ⊂ 𝑅+ ,
forms an increasing sequence converging to infinity (see [11,
Theorem 1.4]). Moreover, there exists an orthonormal basis
in 𝐿2 (Ω) × πΏ2 (Γ), {πœ”π‘› }𝑛 , which are solutions of the eigenvalue
problem
−Δ𝑒 + πœ†π‘’ = πœ‡π‘› 𝑒,
πœ•π‘’
= πœ‡π‘› 𝑒,
πœ•]
π‘₯ ∈ Ω,
(17)
π‘₯ ∈ Γ,
+
where πœ‡π‘› ⊂ 𝑅 forms an increasing sequence converging to
infinity. We denote by 𝑃𝑛 the orthonormal projector
2
2
𝑃𝑛 : 𝐿 (Ω) × πΏ (Γ) 󳨀→ span {πœ”1 , πœ”2 , . . . , πœ”π‘› } ,
∞
𝑛
𝑖=1
𝑖=1
(18)
𝑒 = ∑𝛼𝑖 πœ”π‘– 󳨃󳨀→ 𝑒𝑛 = ∑𝛼𝑖 πœ”π‘– .
So, we can perform the Galerkin truncation by using
orthonormal basis mentioned above and guarantee the following existence and uniqueness (see [1, 8, 9, 12]).
By the last theorem, we can define the operator semigroup
{𝑆(𝑑)}𝑑≥0 in 𝐿2 (Ω) × πΏ2 (Γ) as follows:
𝑆 (𝑑) (𝑒0 , 𝛾 (𝑒0 )) : 𝐿2 (Ω) × πΏ2 (Γ) × R+ 󳨀→ 𝐿2 (Ω) × πΏ2 (Γ) ,
(22)
which is continuous in 𝐿2 (Ω) × πΏ2 (Γ).
Furthermore, we obtained the bounded absorbing set for
dynamical system (𝑆(𝑑), 𝑋) (see [8, 12]).
Theorem 11. Under the assumptions of Theorem 10, the semigroup {𝑆(𝑑)}𝑑≥0 have (𝐿2 (Ω) × πΏ2 (Γ), 𝐿𝑝 (Ω) × πΏπ‘ž (Γ))-, (𝐿2 (Ω) ×
𝐿2 (Γ), 𝐻1 (Ω) × πΏπ‘ž (Γ))-bounded absorbing sets, that is, for any
bounded subset 𝐡 ⊂ 𝐿2 (Ω) × πΏ2 (Γ), there exists a positive
constant 𝑇, which is only dependent on the 𝐿2 (Ω)×𝐿2 (Γ)-norm
of 𝐡, such that
‖𝑒(𝑑)‖𝐿𝑝 (Ω) ≤ 𝑅0
for any (𝑒0 , 𝛾 (𝑒0 )) ∈ 𝐡, 𝑑 ≥ 𝑇,
‖𝑒(𝑑)β€–πΏπ‘ž (Γ) ≤ 𝑅0
for any (𝑒0 , 𝛾 (𝑒0 )) ∈ 𝐡, 𝑑 ≥ 𝑇,
β€–∇𝑒(𝑑)‖𝐿2 (Ω) ≤ 𝑅0
𝑝 ≥ 2,
−𝐢3 + 𝐢4 |𝑠|π‘ž ≤ 𝑔 (𝑠) 𝑠 ≤ 𝐢3 + 𝐢5 |𝑠|π‘ž ,
π‘ž ≥ 2,
𝑓󸀠 (𝑠) ≥ −𝑙,
𝑔󸀠 (𝑠) ≥ −π‘š.
3. The Main Results
3.1. Exponential Attractors in 𝐿2 (Ω) × πΏ2 (Γ)
Theorem 12. Under the hypothesis of Theorem 11, {𝑆(𝑑)}𝑑≥0 has
a (𝐿2 (Ω) × πΏ2 (Γ), 𝐿2 (Ω) × πΏ2 (Γ))-exponential attractor.
Proof. Let 𝑒2 = (𝐼 − 𝑃𝑛 )𝑒, where 𝑃𝑛 is denoted by the
orthonormal projector as mentioned before. Multiplying (1)1
by 𝑒2 and integrating by parts, we get
1 𝑑 σ΅„©σ΅„© σ΅„©σ΅„©2 σ΅„©σ΅„© σ΅„©σ΅„©2
(󡄩𝑒 σ΅„© + 󡄩𝑒 σ΅„© )
2 𝑑𝑑 σ΅„© 2 σ΅„©Ω σ΅„© 2 σ΅„©Γ
󡄨
󡄨2
+ ∫ 󡄨󡄨󡄨∇𝑒2 󡄨󡄨󡄨 + ∫ 𝑓 (𝑒) 𝑒2 + ∫ 𝑔 (𝑒) 𝑒2
Ω
Ω
Γ
= (β„Ž, 𝑒2 ) + ⟨𝜌, 𝑒2 ⟩ .
(19)
(20)
𝑒 ∈ C ([0, 𝑇] ; 𝐿2 (Ω)) ,
𝑒 ∈ C ([0, 𝑇] ; 𝐿2 (Γ)) ,
(24)
Using (19), we can deduce that
Then, Problem (1) has a unique weak solution, for any 𝑇 > 0,
given (𝑒0 , 𝛾(𝑒0 )) ∈ 𝐿2 (Ω) × πΏ2 (Γ) and β„Ž(π‘₯) ∈ 𝐿2 (Ω), 𝜌(π‘₯) ∈
𝐿2 (Γ) there exists a solution 𝑒 with
𝑒 ∈ 𝐿2 (0, 𝑇; 𝐻1 (Ω)) ∩ 𝐿𝑝 (Ω𝑇 ) ,
for any (𝑒0 , 𝛾 (𝑒0 )) ∈ 𝐡, 𝑑 ≥ 𝑇,
where 𝑅0 is a positive constant independent of 𝐡, 𝑒(𝑑) = 𝑆(𝑑)𝑒0 .
Theorem 10. Assume that the functions 𝑓, 𝑔 ∈ C1 satisfying
−𝐢0 + 𝐢1 |𝑠|𝑝 ≤ 𝑓 (𝑠) 𝑠 ≤ 𝐢0 + 𝐢2 |𝑠|𝑝 ,
(23)
(21)
𝑒 ∈ πΏπ‘ž (Γ𝑇 ) ,
and (𝑒0 , 𝛾(𝑒0 )) 󳨃→ (𝑒(𝑑), 𝛾(𝑒)) is continuous on 𝐿2 (Ω) × πΏ2 (Γ).
|𝑒|𝑝 ≥ |𝑒|2 − 𝐢,
|𝑒|π‘ž ≥ |𝑒|2 − 𝐢,
(25)
for any 𝑝, π‘ž ≥ 2. Therefore,
∫ 𝑓 (𝑒) 𝑒2 + ∫ 𝑔 (𝑒) 𝑒2
Ω
Γ
󡄨 󡄨𝑝
󡄨 σ΅„¨π‘ž
≥ ∫ 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨 + ∫ 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨 − 𝐢 (π‘š (Ω) , π‘š (Γ)) .
Ω
Γ
(26)
Thus, we know
𝑑 σ΅„©σ΅„© σ΅„©σ΅„©2 σ΅„©σ΅„© σ΅„©σ΅„©2
󡄨 󡄨2
(󡄩𝑒 σ΅„© + 󡄩𝑒 σ΅„© ) + πœ‡π‘› (∫ 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨 + ∫ |𝑒|2 )
𝑑𝑑 σ΅„© 2 σ΅„©Ω σ΅„© 2 σ΅„©Γ
Ω
Γ
󡄨 󡄨󡄨
󡄨
≤ 𝐢 (π‘š (Ω) , π‘š (Γ) , |β„Ž|Ω , σ΅„¨σ΅„¨πœŒσ΅„¨σ΅„¨Γ ) .
(27)
Journal of Applied Mathematics
5
Applying the Gronwall-inequality, we have
𝐢
σ΅„©σ΅„©
σ΅„©2 σ΅„©
σ΅„©2
−πœ‡ 𝑑
󡄩󡄩𝑒2 (𝑑)σ΅„©σ΅„©σ΅„©Ω + 󡄩󡄩󡄩𝑒2 (𝑑)σ΅„©σ΅„©σ΅„©Γ ≤ 𝑒 𝑛 + .
πœ‡π‘›
Applying Theorem 9 to Theorems 11 and 16, we have the
following.
(28)
Obviously, the enhanced flattening property holds true. By
Theorem 6, we obtain the exponential attractor in 𝐿2 (Ω) ×
𝐿2 (Γ).
Corollary 17. Under the assumption of Theorem 12, the semigroup {𝑆(𝑑)}𝑑≥0 generated by Problem (1) with initial data 𝑒0 ∈
𝐿2 (Ω) and 𝛾(𝑒0 ) ∈ 𝐿2 (Γ) has a (𝐿2 (Ω)×𝐿2 (Γ), 𝐿𝑝 (Ω)×πΏπ‘ž (Γ))exponential attractor Mπ‘π‘ž .
3.2. (𝐿2 (Ω) × πΏ2 (Γ),πΏπ‘Ÿ (Ω) × πΏπ‘Ÿ (Γ))-Exponential Attractor. If
π‘Ÿ = min{𝑝, π‘ž}, we easily showed asymptotic a prior estimate
of the solution of (1) in πΏπ‘Ÿ (Ω)×πΏπ‘Ÿ (Γ) in order to obtain global
and uniform attractors in 𝐿𝑝 (Ω) × πΏπ‘ (Γ), where π‘Ÿ(≥ 2) is an
integer (resp., see [8, 12]).
Remark 18. In fact, for any integer π‘Ÿ ≤ max{𝑝, π‘ž}, we can
obtain Theorem 16 and Corollary 17 by Lemma 15. In other
words, if π‘Ÿ ≤ min{𝑝, π‘ž}, we can obtain Theorem 16 and
Corollary 17 by Lemma 15 and need not use asymptotic a
prior estimate. Here, we point this result obtained by the
different procedure.
Theorem 13. Under the hypothesis of Theorem 11, then for any
πœ€ > 0 and any bounded subset 𝐡 ⊂ 𝐿2 (Ω) × πΏ2 (Γ), there exist
two positive constants 𝑇 = 𝑇(𝐡, πœ€) and 𝑀 = 𝑀(πœ€) such that
On account of Theorem 6 and Corollary 17, we have the
following.
∫
Ω(|𝑒|≥𝑀)
π‘Ÿ
|𝑒| + ∫
Γ(|𝑒|≥𝑀)
π‘Ÿ
|𝑒| ≤ πΆπœ€,
Corollary 19. There exists some π‘š such that
∀𝑑 ≥ 𝑇,
(29)
(𝑒0 , 𝛾 (𝑒0 )) ∈ 𝐡,
σ΅„©
σ΅„©
σ΅„©
σ΅„©σ΅„©
−𝑙𝑑
σ΅„©σ΅„©(𝐼 − π‘ƒπ‘š )𝑒󡄩󡄩󡄩𝐿𝑝 (Ω) + σ΅„©σ΅„©σ΅„©(𝐼 − π‘ƒπ‘š )π‘’σ΅„©σ΅„©σ΅„©πΏπ‘ž (Γ) ≤ π‘˜π‘’ + π‘˜ (π‘š) .
(31)
where the constant 𝐢 is independent of πœ€ and 𝐡.
Now, we give some a prior estimates about 𝑒𝑑 .
After Theorem 13, we obtain the subsequent result.
Theorem 14. Under the assumption of Theorem 11, then the
semigroup {𝑆(𝑑)}𝑑≥0 generated by Problem (1) with initial data
𝑒0 ∈ 𝐿2 (Ω) and 𝛾(𝑒0 ) ∈ 𝐿2 (Γ) has a (𝐿2 (Ω) × πΏ2 (Γ), πΏπ‘Ÿ (Ω) ×
πΏπ‘Ÿ (Γ))-exponential attractor Mπ‘Ÿ , that is, Mπ‘Ÿ is compact,
invariant in πΏπ‘Ÿ (Ω) × πΏπ‘Ÿ (Γ), and attracts every bounded in
𝐿2 (Ω) × πΏ2 (Γ) in the topology of πΏπ‘Ÿ (Ω) × πΏπ‘Ÿ (Γ).
Proof. By Theorems 11 and 12, it is easily verified that {𝑆(𝑑)}𝑑≥0
has an exponential attractor in πΏπ‘Ÿ (Ω) × πΏπ‘Ÿ (Γ) where we apply
Theorem 9.
Lemma 20 (see [8, 12]). Under the assumption of Theorem 11,
for any bounded subset 𝐡 ⊂ 𝐿2 (Ω) × πΏ2 (Γ), there exists a
positive constant 𝑇1 which depends only on the 𝐿2 (Ω) × πΏ2 (Γ)norm of B such that
󡄨2
󡄨2
󡄨
󡄨
∫ 󡄨󡄨󡄨𝑒𝑑 (𝑠)󡄨󡄨󡄨 𝑑π‘₯ + ∫ 󡄨󡄨󡄨𝑒𝑑 (𝑠)󡄨󡄨󡄨 𝑑π‘₯ ≤ 𝑅1
Ω
Γ
∀𝑠 ≥ 𝑇1 , (𝑒0,0 ) ∈ 𝐡,
(32)
If π‘Ÿ > min{𝑝, π‘ž}, the reasoning process mentioned earlier
is not available. We note the following result in [13] after
authors obtained regularity of global attractor.
where 𝑅1 is a positive constant which depends on 𝑀.
Lemma 15 (see [13, Corollary 3.1]). Under the assumptions
of Theorem 11, the semigroup {𝑆(𝑑)}𝑑≥0 has a compact global
attractor A in (𝐻1 (Ω) ∩ 𝐿𝑝 (Ω)) × πΏπ‘ž (Γ).
Theorem 21. Under the assumption of Theorem 11, the semigroup {𝑆(𝑑)}𝑑≥0 generated by Problem (1) with initial data 𝑒0 ∈
𝐿2 (Ω) and 𝛾(𝑒0 ) ∈ 𝐿2 (Γ) satisfies the enhanced flattering in the
space 𝐻1 (Ω) × πΏπ‘ž (Γ).
By Theorem 7 and Lemma 15, we easily obtain the following.
Theorem 16. Under the hypothesis of Theorem 11, then for any
πœ€ > 0 and any bounded subset 𝐡 ⊂ 𝐿2 (Ω) × πΏ2 (Γ), there exist
two positive constants 𝑇 = 𝑇(𝐡, πœ€) and 𝑀 = 𝑀(πœ€) such that
∫
Ω(|𝑒|≥𝑀)
𝑝
|𝑒| + ∫
Γ(|𝑒|≥𝑀)
π‘ž
|𝑒| ≤ πΆπœ€,
(30)
where the constant 𝐢 is independent of πœ€ and 𝐡.
Proof. We denote 𝑒 = 𝑒1 + 𝑒2 , where 𝑒1 = 𝑃𝑛 𝑒, 𝑒2 = (𝐼−𝑃𝑛 )𝑒,
and 𝑃𝑛 has been introduced in Section 3.
Multiplying Problem (1) by 𝑒2 , we can get
1 𝑑
󡄨 󡄨2 1 𝑑
󡄨 󡄨2
󡄨
󡄨2
∫ 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨 +
∫ 󡄨󡄨𝑒 󡄨󡄨 + ∫ 󡄨󡄨󡄨∇𝑒2 󡄨󡄨󡄨
2 𝑑𝑑 Ω
2 𝑑𝑑 ٠󡄨 2 󡄨
Ω
∀𝑑 ≥ 𝑇,
(𝑒0 , 𝛾 (𝑒0 )) ∈ 𝐡,
We can easily obtain the following theorem.
+ ∫ 𝑓 (𝑒) 𝑒2 + ∫ 𝑔 (𝑒) 𝑒2
Ω
Γ
= (β„Ž, 𝑒2 ) + ⟨𝜌, 𝑒2 ⟩ .
(33)
6
Journal of Applied Mathematics
By (25) and (26), we can deduce
󡄨
󡄨2
󡄨 󡄨
󡄨 󡄨
∫ 󡄨󡄨󡄨∇𝑒2 󡄨󡄨󡄨 ≤ (󡄨󡄨󡄨𝑒𝑑2 󡄨󡄨󡄨٠+ |β„Ž|Ω ) 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨Ω
Ω
󡄨 󡄨 󡄨 󡄨 󡄨 󡄨
+ (󡄨󡄨󡄨𝑒𝑑2 󡄨󡄨󡄨à + σ΅„¨σ΅„¨σ΅„¨πœŒσ΅„¨σ΅„¨σ΅„¨Γ ) 󡄨󡄨󡄨𝑒2 󡄨󡄨󡄨Γ
𝑝−1 σ΅„© σ΅„©
+ ‖𝑒‖٠󡄩󡄩󡄩𝑒2 󡄩󡄩󡄩𝐿𝑝 (Ω)
π‘ž−1 σ΅„© σ΅„©
+ ‖𝑒‖à 󡄩󡄩󡄩𝑒2 σ΅„©σ΅„©σ΅„©πΏπ‘ž (Γ) .
(34)
By Theorems 11 and 12, Corollary 19, and Lemma 20, we
know that 𝑆(𝑑)𝑒0 satisfies the enhanced flattering property in
𝐻1 (Ω). On account of Theorem 6 and Corollary 19, we can
deduce that Theorem 21 is valid.
It is immediate by Theorem 6, Corollary 19, and
Theorem 21.
Corollary 22. Under the assumption of Theorem 12, the semigroup {𝑆(𝑑)}𝑑≥0 generated by Problem (1) with initial data
(𝑒0 , 𝛾(𝑒0 )) ∈ 𝐿2 (Ω) × πΏ2 (Γ) has a (𝐿2 (Ω) × πΏ2 (Γ), (𝐻1 (Ω) ∩
𝐿𝑝 (Ω)) × πΏπ‘ž (Γ))-exponential attractor M.
Acknowledgment
The research was supported by the NSFC Grant (no.
11271339).
References
[1] M. Anguiano, P. Marı́n-Rubio, and J. Real, “Pullback attractors
for non-autonomous reaction-diffusion equations with dynamical boundary conditions,” Journal of Mathematical Analysis and
Applications, vol. 383, no. 2, pp. 608–618, 2011.
[2] J. M. Arrieta, P. Quittner, and A. Rodrı́guez-Bernal, “Parabolic
problems with nonlinear dynamical boundary conditions and
singular initial data,” Differential and Integral Equations, vol. 14,
no. 12, pp. 1487–1510, 2001.
[3] A. Rodrı́guez-Bernal, “Attractors for parabolic equations with
nonlinear boundary conditions, critical exponents, and singular
initial data,” Journal of Differential Equations, vol. 181, no. 1, pp.
165–196, 2002.
[4] A. Constantin and J. Escher, “Global solutions for quasilinear
parabolic problems,” Journal of Evolution Equations, vol. 2, no.
1, pp. 97–111, 2002.
[5] A. Constantin and J. Escher, “Global existence for fully parabolic
boundary value problems,” Nonlinear Differential Equations and
Applications, vol. 13, no. 1, pp. 91–118, 2006.
[6] A. Constantin, J. Escher, and Z. Yin, “Global solutions for
quasilinear parabolic systems,” Journal of Differential Equations,
vol. 197, no. 1, pp. 73–84, 2004.
[7] J. Escher, “Quasilinear parabolic systems with dynamical
boundary conditions,” Communications in Partial Differential
Equations, vol. 18, no. 7-8, pp. 1309–1364, 1993.
[8] Z. H. Fan and C. K. Zhong, “Attractors for parabolic equations
with dynamic boundary conditions,” Nonlinear Analysis: Theory, Methods & Applications, vol. 68, no. 6, pp. 1723–1732, 2008.
[9] C. G. Gal and M. Warma, “Well posedness and the global
attractor of some quasi-linear parabolic equations with nonlinear dynamic boundary conditions,” Differential and Integral
Equations, vol. 23, no. 3-4, pp. 327–358, 2010.
[10] J. Petersson, “A note on quenching for parabolic equations with
dynamic boundary conditions,” Nonlinear Analysis: Theory,
Methods & Applications, vol. 58, no. 3-4, pp. 417–423, 2004.
[11] L. Popescu and A. Rodriguez-Bernal, “On a singularly perturbed wave equation with dynamic boundary conditions,”
Proceedings of the Royal Society of Edinburgh A, vol. 134, no. 2,
pp. 389–413, 2004.
[12] Y. Lu, “Uniform attractors for the closed process and applications to the reaction-diffusion equation with dynamical
boundary condition,” Nonlinear Analysis: Theory, Methods &
Applications, vol. 71, no. 9, pp. 4012–4025, 2009.
[13] L. Yang and M. Yang, “Long-time behavior of reaction-diffusion
equations with dynamical boundary condition,” Nonlinear
Analysis: Theory, Methods & Applications, vol. 74, no. 12, pp.
3876–3883, 2011.
[14] Y. Li, H. Wu, and T. Zhao, “Necessary and sufficient conditions
for the existence of exponential attractors for semigroups, and
applications,” Nonlinear Analysis: Theory, Methods & Applications, vol. 75, no. 17, pp. 6297–6305, 2012.
[15] C. K. Zhong, M. H. Yang, and C. Y. Sun, “The existence of
global attractors for the norm-to-weak continuous semigroup
and application to the nonlinear reaction-diffusion equations,”
Journal of Differential Equations, vol. 223, no. 2, pp. 367–399,
2006.
[16] Y. Zhong and C. Zhong, “Exponential attractors for reactiondiffusion equations with arbitrary polynomial growth,” Nonlinear Analysis: Theory, Methods & Applications, vol. 71, no. 3-4, pp.
751–765, 2009.
[17] L. Dung and B. Nicolaenko, “Exponential attractors in Banach
spaces,” Journal of Dynamics and Differential Equations, vol. 13,
no. 4, pp. 791–806, 2001.
[18] A. Eden, C. Foias, B. Nicolaenko, and R. Temam, Exponential
Attractors for Dissipative Evolution Equations, vol. 37 of Research
in Applied Mathematics, John Wiley & Sons, New York, NY,
USA, 1994.
[19] M. Grasselli and D. Pražák, “Exponential attractors for a class
of reaction-diffusion problems with time delays,” Journal of
Evolution Equations, vol. 7, no. 4, pp. 649–667, 2007.
[20] R. Temam, Infinite-Dimensional Dynamical Systems in Mechanics and Physics, vol. 68 of Applied Mathematical Sciences,
Springer, New York, 1997.
[21] Q. Ma, S. Wang, and C. Zhong, “Necessary and sufficient
conditions for the existence of global attractors for semigroups
and applications,” Indiana University Mathematics Journal, vol.
51, no. 6, pp. 1541–1559, 2002.
Advances in
Operations Research
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Advances in
Decision Sciences
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Mathematical Problems
in Engineering
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Journal of
Algebra
Hindawi Publishing Corporation
http://www.hindawi.com
Probability and Statistics
Volume 2014
The Scientific
World Journal
Hindawi Publishing Corporation
http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
International Journal of
Differential Equations
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Volume 2014
Submit your manuscripts at
http://www.hindawi.com
International Journal of
Advances in
Combinatorics
Hindawi Publishing Corporation
http://www.hindawi.com
Mathematical Physics
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Journal of
Complex Analysis
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
International
Journal of
Mathematics and
Mathematical
Sciences
Journal of
Hindawi Publishing Corporation
http://www.hindawi.com
Stochastic Analysis
Abstract and
Applied Analysis
Hindawi Publishing Corporation
http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com
International Journal of
Mathematics
Volume 2014
Volume 2014
Discrete Dynamics in
Nature and Society
Volume 2014
Volume 2014
Journal of
Journal of
Discrete Mathematics
Journal of
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Applied Mathematics
Journal of
Function Spaces
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Optimization
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Download