R E S E A R C H
Open Access
Averaging of the
D
non-autonomous
Benjamin-Bona-Mahony equation with
singularly oscillating forces
Mingxia Zhao
1, Xinguang Yang
2*and Lingrui Zhang
3*Correspondence:
2College of Mathematics and
Information Science, Henan Normal University, Xinxiang, 453007, P.R. China
Full list of author information is available at the end of the article
Abstract
For
ε
∈(0, 1), we investigate the convergence of corresponding uniform attractors of the 3Dnon-autonomous Benjamin-Bona-Mahony equation with singularly oscillating force contrast with the averaged Benjamin-Bona-Mahony equation (corresponding to the limiting caseε
= 0). Under suitable assumptions on the external force, we shall obtain the uniform boundedness and convergence of the related uniform attractors asε
→0+.MSC: 35B40; 35Q99; 80A22
Keywords: Benjamin-Bona-Mahony equation; singularly oscillating forces; uniform attractors; translational bounded functions
1 Introduction
Let ρ ∈[, ) be a fixed parameter, ⊂ R be a bounded domain with sufficiently smooth boundary∂. We investigate the long-time behavior for the non-autonomous DBenjamin-Bona-Mahony (BBM) equation with singularly oscillating forces:
ut–ut–νu+∇ ·
–
→
F(u) =f(t,x) +ε–ρf(t/ε,x), x∈, (.)
u(t,x)|∂= , (.)
u(τ,x) =uτ(x), τ∈R. (.)
Here,t∈Rτ,Rτ = (τ,∞), andu=u(t,x) = (u(t,x),u(t,x),u(t,x)) is the velocity vector
field,ν> is the kinematic viscosity,→–F is a nonlinear vector function,f(t,x) +ε–ρf(t/ε,x)
is the singularly oscillating force.
Along with (.)-(.), we consider the averaged Benjamin-Bona-Mahony equation
ut–ut–νu+∇ ·
–
→
F(u) =f(t,x), x∈, (.)
u(t,x)|∂= , (.)
u(τ,x) =uτ(x), τ∈R (.)
formally corresponding to the caseε= in (.).
The function
fε(x,t) =
f(x,t) +ε–ρf(x,t/ε), <ε< ,
f(x,t), ε=
(.)
represents the external forces of problem (.)-(.) forε> and of problem (.)-(.) for ε= , respectively.
The functions f(x,s) and f(x,s) are taken from the space Lb(R,H) of translational
bounded functions inL
loc(R,H), namely,
fL b(R,H)
:=sup
t∈R
t+
t
f(s)
Hds=M
, (.)
fL b(R,H):=
sup
t∈R
t+
t
f(s)
Hds=M
, (.)
for some constantsM,M≥.
Defining
Qε=
M+ Mε–ρ, <ε< ,
M, ε= ,
as a straightforward consequence of (.), we have
fεL
b(R,H)≤Q ε
, (.)
note thatQεis of the orderε–ρasε→+.
The BBM equation is a well-known model for long waves in shallow water which was in-troduced by Benjamin, Bona, and Mahony ([], ) as an improvement of the Korteweg-de Vries equation (KdV equation) for moKorteweg-deling long waves of small amplituKorteweg-de in two dimensions. Contrasting with the KdV equation, the BBM equation is unstable in high wavenumber components. Further, while the KdV equation has an infinite number of in-tegrals of motion, the BBM equation only has three. For more results on the wellposedness and infinite dimensional dynamical systems for BBM equations, we can refer to [–].
In this paper, firstly, we shall study the asymptotic behavior of the non-autonomous BBM equation depending on the small parameterε, which reflects the rate of fast time oscil-lations in the termε–ρf
(x,t/ε) with amplitude of orderε–ρ, then we shall consider the
boundedness and convergence of corresponding uniform attractors of (.)-(.) in con-trast to (.)-(.).
2 Preliminaries
Throughout this paper,Lp() (≤p≤+∞) is the generic Lebesgue space,Hs() is the
Sobolev space. We setE:={u|u∈(C∞())},H,V,W is the closure of the setEin the
topology of (L()), (H()), (H())respectively. ‘’ stands for the weak convergence
Lemma . For eachτ ∈R,every nonnegative locally summable functionφ on Rτ and
everyβ> ,we have
t
τ
φ(s)e–β(t–s)ds≤
–e–β sup θ≥τ
θ+
θ
φ(s)ds, (.)
holds for all t≥τ.
Proof See,e.g., [].
Lemma . Letζ:Rτ →R+fulfill that for almost every t≥τ,the differential inequality
d
dtζ(t) +φ(t)ζ(t)≤φ(t), (.)
where,for every t≥τ,the scalar functionsφandφsatisfy
t
τ
φ(s)ds≥β(t–τ) –γ,
t+
t
φ(s)ds≤M, (.)
for someβ> ,γ ≥and M≥.Then
ζ(t)≤eγζ(τ)e–β(t–τ)+ Me
γ
–e–β, ∀t≥τ. (.)
Proof See,e.g., [].
For the non-autonomous general Benjamin-Bona-Mahony (BBM) equation,
ut–ut–νu+∇ ·
–
→
F(u) =g(t,x), x∈,t∈Rτ, (.)
u(t,x)|∂= , (.)
u(τ,x) =uτ(x), τ∈R. (.)
Assume thatuτ∈H(), the nonlinear vector function
–
→
F(s) = (F(s),F(s),F(s)),∀s∈R,
we denote
fi(s) =Fi(s), Fi(s) =
s
Fi(r)dr, (.)
where
–
→
f (s) =f(s),f(s),f(s)
, →–F(s) =F(s),F(s),F(s)
. (.)
In addition,Fi(i= , , ) is a smooth function satisfying
Fi() = , Fi(s)≤C|s|+C|s|, (.)
C+C|s| ≤fi(s)≤C+C|s|, Fi(s)≤C|s|+C|s| (.)
Similar to [], by the Galerkin method anda prioriestimate, we easily derive the exis-tence of a global weak solution and a uniform attractor which shall be stated in the follow-ing theorems.
Theorem . Assume that(.)-(.)hold,g∈L
loc(R,H),uτ∈H() (or V) ,then there
exists a unique global weak solution u(x,t)of the problem(.)-(.)which satisfies
u∈C(τ,T);V, ut∈L
(τ,T);V (.)
for allτ∈R and T>τ.
Theorem . Assume that the external force g∈L
loc(R,H)and(.)-(.)hold,then the
processes{U(t,τ),t≥τ}generated by the global solution possess uniform attractorsAg(t)
in H()for the non-autonomous system(.)-(.).
3 Some lemmas
Lemma . The functions f(x,s)and f(x,s)are taken from the space Lb(R,H)of
transla-tional bounded functions in Lloc(R,H),then the processes{Ufε(t,τ),t≥τ,t,τ∈R}
gener-ated by system(.)-(.)have a uniformly(w.r.t.σ=fε∈)compact attractorAεfor any
fixedε∈(, ).
Proof As a similar argument in Section , we chooseg(t,x) =fε(t,x) in Theorem ., since
f andf are translational bounded inLloc(R,H), then for any fixedε∈(, ],fε(t,x) is
translational bounded inL
loc(R,H) and we can easily deduce the existence of uniformly
compact attractorsAε.
We can briefly describe the structure of the uniform attractor as follows: if the func-tionsf(t) andf(t) are translational bounded, problem (.)-(.) generates the dynamical
processes{Uε(t,τ),t≥τ,τ ∈R}acting onVwhich is defined byUε(t,τ)uε τ =u
ε(t),t≥τ,
whereuε(t) is the solution to (.)-(.). The processes{Uε(t,τ),t≥τ,τ∈R}have a
uni-formly (w.r.t.t∈R) absorbing set Bε:=uε∈V|uε
V≤CQ
ε , (.)
which is bounded inV for any fixedε∈(, ).
On the other hand,Aεis also bounded inVfor each fixedεsinceAε⊆Bε
. Assuming
f,f∈Ltc(R,H), the external forcefε(t) appearing in equation (.) belongs toLtc(R,H)
also. Moreover, ifε> andfˆε∈H(fε), then
ˆ
fε(t) =fˆ(t) +ε–ρfˆ
t ε
, (.)
for somefˆ∈H(f) andfˆ∈H(f). In this case, to describe the structure of the uniform
attractorAε, we consider the family of equations
ˆ
ut+Auˆt+νAuˆ+∇ ·F(u) =ˆ fˆε(t), fˆε∈H
fε. (.)
For every external forceˆfε∈H(fε), equation (.) generates a class of processes{U ˆ fε(t,τ)}
to the original equation (.) with the external forcefε(t). Moreover, the map
uτ,fˆε
→Ufˆε(t,τ)uτ (.)
is (V×H(fε),V)-continuous.
Lemma . If the function f(t,x)in(.)is taken from the space Lb(R,H)of translational
bounded functions in L
loc(R,H),then the processes{Uf(t,τ),t≥τ,τ∈R}generated by
sys-tem(.)-(.)have a uniformly(w.r.t.σ=f∈)compact attractorA.
Proof Use a similar technique as that in Theorem ., we can easily deduce the existence of a uniformly compact attractorAif we chooseg(t,x) =f
(t,x).
4 Uniform boundedness ofAε
Firstly, we shall consider the auxiliary linear equation with a non-autonomous external force and give some useful lemmas, and then we shall prove the uniform boundedness ofAε.
Considering the linear equation
Yt+AYt+νAY=K(t), Y|t=τ= , (.)
we get the following lemma.
Lemma . Assume that K∈L
loc(R,H),then problem(.)has a unique solution
Y∈L(τ,T);W∩C(τ,T);V, (.)
∂tY∈L
(τ,T);W. (.)
Moreover,the following inequalities
Y(t)W≤C t
τ
e–Cν(t–s)K(s)Hds, (.)
t+
t
Y(s)Vds≤CY(t)V+ t+
t
K(s)Hds
(.)
hold for every t≥τ and some constant C> ,independent of the initial timeτ∈R. Proof Firstly, using the Galerkin approximation method, we can deduce the existence of a global solution for (.), here we omit the details.
Then multiplying (.) byYandAYrespectively, we get
d dt
Y+∇Y+ν∇Y=K(t),Y≤ νK(t)
+ν Y
(.)
and
d dt
∇Y+AY+νAY=K(t),AY≤ νK(t)
+ν AY
. (.)
SettingK(t,τ) =τtk(s)ds,t≥τ,τ∈R, we have the following lemma. Lemma . Assume that the formula
sup
t≥τ,τ∈R
K(t,τ)H+ t+
t
K(s,τ)Hds
≤l (.)
holds for some constant l≥,let k∈L
loc(R,H).Then the solution y(t)yields the following
problem:
yt+Ayt+νAy=k(t/ε), y|t=τ= , (.)
withε∈(, )satisfying the inequality y(t)V+
t+
t
y(s)Vds≤Clε, ∀t≥τ, (.)
where C> is constant independent of K. Moreover,we also have
t+
t
Kε(s)
Hds≤C. (.)
Proof Noting that
Kε(t) =
t
τ
k(s/ε)ds=ε t/ε
τ/ε
k(s)ds=εK(t/ε,τ/ε), (.)
we can derive the following estimates from (.):
sup
t≥τ
Kε(t)H≤lε,
t+
t
Kε(s)
Hds=ε
t+
t
K(s/ε,τ/ε)Hds
≤Cεsup
t≥τ
t+
t
K(s,τ)Hds
≤Clε.
From Lemma ., we have t
τ
e–Cν(t–s)Kε(s)
Hds
≤
t
t–
eCν(s–t)Kε(s)
ds+ t–
t–
eCν(s–t)Kε(s)
ds+· · ·
≤
t
t–
Kε(s)
ds+e–Cν t–
t–
Kε(s)
ds+e–Cν t–
t–
Kε(s)
ds+· · ·
≤ +e–Cν+e–Cν+· · ·Kε(s)
Lb(R;H)
≤
( –e–Cν)Kε(s)
≤
( –e–Cν)sup t≥τ
t+
t
Kε(s)Hds
≤Clε. (.)
Hence, from the Poincaré inequality, combining (.) and (.)-(.), we conclude that
Y(t)W≤Clε, (.)
t+
t
Y(s)Vds≤CY(t)V+ t+
t
K(s)Hds
≤Clε. (.)
Setting
Y(t) = t
τ
y(s)ds, (.)
we deduce that for anyt≥τ,
∂tY(t) =y(t) =
t
τ
∂ty(s)ds, (.)
sincey(τ) = .
Integrating (.) with respect to time variable fromτtot, we see thatY(t) is a solution to the problem
∂tY(t) +∂t
AY(t)+νAY(t) =Kε(t), qY(t)|t=τ= , (.)
such that from (.) and (.), we can derive
Y(t)H+∇Y(t)H+ t+
t
Y(s)Vds≤Clε. (.)
By virtue ofy(t) =∂tY(t), (AY(t),Y(t))∼ Y(t)V,AY(t) ∼ Y(t)W, we have
∂tY(t)+∂tAY(t)=y(t)+Ay(t)≤νY(t)W+Kε(t)≤Clε. (.)
Hence, we conclude
y(t)V≤Cy(t)+Ay(t)≤CνY(t)W+Kε(t)≤Clε (.)
and t+
t
y(s)Vds≤Clε. (.)
The proof is finished.
Now, we shall use the auxiliary linear equation and some estimates to prove the uniform boundedness ofAεinV. For convenience, we set
F(t,τ) =
t
τ
and assume that
sup
t≥τ,τ∈R
F(t,τ)
+ t+
t
F(s,τ)
Hds
≤l, (.)
holds for some constantsl≥.
Theorem . The attractorsAεof problem(.)-(.) (or(.)-(.))are uniformly(w.r.t.
ε)bounded in V,namely,
sup
ε∈[,)
Aε
V< +∞. (.)
Proof Letuε(t) =Uε(t,τ)uε
τ be the solution to (.)-(.) with the initial datauετ ∈V. For
ε> , we consider the auxiliary linear equation
vt+Avt+νAv=ε–ρf(t/ε), v|t=τ = . (.)
From Lemma ., we have the estimate
v(t)V+ t+
t
v(s)Vds≤Clε(–ρ), ∀t≥τ. (.)
Setting the functionw(t) as
w(t) =u(t) –v(t), (.)
which satisfies the problem
wt+Awt+νAw+∇ ·
–
→
F(w+v) =f, w|t=τ=uτ. (.)
Taking the scalar product of (.) withw, we obtain
d dt
w+∇w+ν∇w+∇ ·→–F(w+v),w= (f,w). (.)
Using the inequality
v(t)=v(t)H≤Cv(t)V≤Clε(–ρ), ∀t≥τ, (.)
we have
∇ ·–→F(w+v),w≤C
+w+v+ ν λw
≤C
+w+lε(–ρ)+ ν λw
≤C
+w+lε(–ρ)+ ν λw
, (.)
Moreover, notice that
(f,w)≤
ν w
V+
νf
, (.)
and inserting (.)-(.) into (.), we have
d dt
w+∇w+ν∇w
≤C
+w+lε(–ρ)+ ν λw
+ν
w
V+
νf
≤C
+w+lε(–ρ)+ν w
V+
ν w
V+
νf
=C
+w+lε(–ρ)+ν w
V+
νf
, (.)
which implies that
d dt
w+wV+φwV≤φ, (.)
where
φ(t)≡
ν –C
+u+lε(–ρ), (.)
φ(t)≡
νf(t)
. (.)
Therefore using (.), we derive from (.)-(.) that for anyt≥τ, t
τ
φ(s)ds≥
ν
(t–τ), (.)
t+
t
φ(s)ds≤CM. (.)
Applying Lemma . withζ(t) =w+w
V,β= ν
,γ = ,M=CM
, we have
w+wV≤Ce–β(t–τ)uτ+uτV
+CM, ∀t≥τ, (.) which gives
wV≤Ce–β(t–τ)u
τ+uτV
+CM, ∀t≥τ. (.)
Recalling thatu=w+v, and using (.) and (.), we end up with
u(t)V≤ wV +vV≤Ce–β(t–τ)uτ+uτV
+Cl+M, ∀t≥τ. (.) Thus, for every <ε≤ε, the processes{Uε(t,τ)}have an absorbing set
B:=
On the other hand, ifε<ε< , the processes{Uε(t,τ)}also possess an absorbing set
Bε=u∈V|u
V≤CQε . (.)
In conclusion, for everyε∈[, ), the set
B*:=B∪Bε (.)
is an absorbing set for{Uε(t,τ)}which is independent ofε. SinceAε⊂B*, (.) follows
and hence the proof is complete.
5 Convergence ofAεtoA0
The main result of the paper reads as follows.
Theorem . Assume that f,f∈Ltc(R,H)⊂Lb(R,H)and(.)holds.Then the uniform
attractorAε(for problem(.)-(.))converges toA(for problem(.)-(.))asε→+in
the following sense:
lim
ε→+distV
Aε,A= . (.)
Next, we shall study the difference of two solutions for (.) withε> and (.) with ε= which share the same initial data. Denote
uε(t) :=Uε(t,τ)uτ, (.)
withuτ belonging to the absorbing setB*which can be found in Section . In particular,
sinceuτ ∈B*, the formula corresponding toε=
u(t)V+ t+
t
u(s)Vds≤R, (.)
holds for someR=R(ρ), as the size ofB*depends onρ.
Lemma . For everyε∈(, ),τ ∈R,uτ∈B*and uε() =u() =uτ,the difference
w(t) =uε(t) –u(t) (.)
satisfies the estimate
w(t)V≤Dε–ρeR(t–τ), ∀t≥τ, (.)
for some positive constants D=D(ρ,l)and R=R(ρ,l),both independent ofε> .
Proof Since the differencew(t) solves the equation
wt+Awt+νAw+∇ ·
–→
the difference
q(t) =w(t) –v(t), (.)
fulfills the Cauchy problem
qt+Aqt+νAq+∇ ·
–→
Fuε–→–Fu= , q|t=τ = , (.)
wherev(t) is the solution to (.).
Taking an inner product of equation (.) withqinH, we obtain
d dt
q+∇q+ν∇q+∇ ·–→Fuε–→–Fu,q= . (.)
Noting that
∇ ·–→Fuε–→–Fu,q
≤sup
i
Fiuε+F i
u∇ ·uε–∇ ·u
+ ν λq
≤C
+uε+u∇w+ ν λq
≤C
+uε+R
∇w+ ν λq
≤C
+uε+Rq+vV+ ν λq
≤C
+K+RvV+ν q
V+
ν λq
=f(t) +ν q
V+h(t)q, (.)
whereλis the first eigenvalue of –,Kis the upper bound foruε(by Lemma .) and
h(t) = ν λ,
f(t) =C
+K+R
vV≤C +K+R
lε(–ρ),
thus, it follows from (.) and (.) that
d dt
q+∇q+ν q
V≤Ch(t)q+f(t)
≤Ch(t)q+∇q+f(t). (.)
Noting thatq(τ)=q(τ)V= , by the Gronwall inequality, we get
q+∇q≤exp
C t τ h(s)ds t τ
f(s)ds. (.)
Moreover, we can derive the following formulas: t
τ
h(s)ds≤ ν
and t
τ
f(s)ds= t
τ
C +K+RvVds
≤
t
τ
C +K+Rlε(–ρ)ds
=C +K+Rlε(–ρ)(t–τ). (.)
Consequently,
q(t)
V ≤C
q+∇q
≤C +K+Rlε(–ρ)(t–τ+ )eνλ(t–τ+)
≤CDε(–ρ)eνλ(t–τ) (.)
holds for some positive constantsD=D(ρ,l). Finally, sincew=q+v, using (.) to
controlvV, we may obtain
w(t)V≤CqV+vV
≤CDε(–ρ)eνλ(t–τ)+Clε(–ρ)
≤Dε(–ρ)eR(t–τ), (.)
whereRis a positive constant. The proof is finished.
Next, we want to generalize Lemma . to derive the convergence of corresponding uniform attractors. Let the external force in equation (.) asfˆ=ˆfε∈H(fε), thenfˆ
∈H(f)
satisfies inequality (.). Define
ˆ
G(t,τ) =
t
τ
ˆ
f(s)ds, t≥τ, (.)
we have
sup
t≥τ,τ∈R
ˆ G(t,τ)
H+
t+
t
G(s,ˆ τ)Hds
≤l. (.)
For anyε∈[, ], we observe thatuˆε(t) =U ˆ
fε(t,τ)yτ is a solution to (.) with the external
forcefˆε=fˆ
+ε–ρfˆ(·/ε)∈H(fε) andyτ(fε)∈B*. Forε> , we investigate the property of
the difference
ˆ
w(t) =uˆε(t) –uˆ(t). (.)
Lemma . The inequality
w(t)ˆ ≤Dε–ρeR(t–τ), ∀t≥τ, (.)
Proof As the similar discussion in the proof of Lemma ., replacinguˆε,ˆf
andfˆ by
uε, f
andf, respectively, noting that (.) still holds for uˆ, and the family{Uˆfε(t,τ)}
(fˆε∈H(fε)), is (H×Hε(fε),H)-continuous, using (.) in place of (.), we can finally
complete the proof of the lemma.
Proof of Theorem. Forε> ,uε∈Aε, we obtain that there exists a complete bounded
trajectoryuˆε(t) of equation (.), with some external force
ˆ
fε=fˆ+ε–ρfˆ(·/ε)∈H
fε, (.)
such thatuˆε() =uε.
We chooseL≥ such that
ˆ
uε(–L)∈Aε⊂B*. (.)
From the equality
uε=Uˆf(, –L)uˆε(–L), (.)
applying Lemma . witht= ,τ= –L, we obtain
uε–U ˆ
f(, –L)uˆε(–L)V≤Dε–ρeRL. (.)
On the other hand, the setAattracts all setsUfˆ(t, –L)B*uniformly whenfˆ∈H(f).
Then, for allδ> , there exists some timeT=T(δ)≥ which is independent ofLsuch that
distV
Ufˆ(T–L, –L)uˆε(–L),A
≤δ. (.)
ChoosingL=T and collecting (.)-(.), we readily get
distV
uε,A≤uε–Ufˆ(, –T)uˆε(–T)V+distV
Uˆf(, –T)uˆε(–T),A
≤Dε–ρeRT+δ. (.)
Sinceuε∈Aεandδ> is arbitrary, taking the limitε→+, we can prove the theorem.
Competing interests
The authors declare that they have no competing interests. Authors’ contributions
The authors declare that the work was realized in collaboration with the same responsibility. All authors read and approved the final manuscript.
Author details
1College of Mathematics and Information Science, Pingdingshan University, Pingdingshan, 467009, P.R. China.2College of
Mathematics and Information Science, Henan Normal University, Xinxiang, 453007, P.R. China.3College of Education and
Acknowledgements
All authors give their thanks to the reviewer’s suggestions, XY was in part supported by the Innovational Scientists and Technicians Troop Construction Projects of Henan Province (No. 114200510011) and the Young Teacher Research Fund of Henan Normal University (qd12104).
Received: 29 January 2013 Accepted: 16 April 2013 Published: 30 April 2013
References
1. Benjamin, TB, Bona, JL, Mahony, JJ: Model equations for long waves in nonlinear dispersive systems. Philos. Trans. R. Soc. Lond. Ser. A272, 47-48 (1972)
2. Avrin, J, Goldstein, JA: Global existence for the Benjamin-Bona-Mahony equation in arbitrary dimensions. Nonlinear Anal. TMA9(8), 861-865 (1985)
3. Goldstein, JA, Wichnoski, BJ: On the Benjamin-Bona-Mahony equation in higher dimensions. Nonlinear Anal. TMA4, 665-675 (1980)
4. Park, JY, Park, SH: Pullback attractors for the non-autonomous Benjamin-Bona-Mahony equation in bounded domains. Sci. China Math.54(4), 741-752 (2011)
5. Qin, Y, Yang, X, Liu, X: Pullback attractors for the non-autonomous Benjamin-Bona-Mahony equations inH2. Acta.
Math. Sci.32B(4), 1338-1348 (2012)
6. Stanislavova, M, Stefanow, A, Wang, B: Asymptotic smoothing and attractors for the generalized Benjamin-Bona-Mahony equation onRn. J. Differ. Equ.219, 451-483 (2005)
7. Wang, B: Strong attractors for the Benjamin-Bona-Mahony equation. Appl. Math. Lett.10, 23-28 (1997) 8. Chepyzhov, VV, Vishik, MI: Attractors for Equations of Mathematical Physics. Am. Math. Soc., Providence (2001)
doi:10.1186/1687-2770-2013-111