Document 10910893

advertisement
Hindawi Publishing Corporation
International Journal of Stochastic Analysis
Volume 2013, Article ID 868301, 16 pages
http://dx.doi.org/10.1155/2013/868301
Research Article
Measure-Dependent Stochastic Nonlinear Beam Equations
Driven by Fractional Brownian Motion
Mark A. McKibben
Department of Mathematics, West Chester University of Pennsylvania, 25 University Avenue,
West Chester, PA 19343, USA
Correspondence should be addressed to Mark A. McKibben; mmckibben@wcupa.edu
Received 1 October 2013; Accepted 9 November 2013
Academic Editor: Ciprian A. Tudor
Copyright © 2013 Mark A. McKibben. 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 a class of nonlinear stochastic partial differential equations arising in the mathematical modeling of the transverse motion
of an extensible beam in the plane. Nonlinear forcing terms of functional-type and those dependent upon a family of probability
measures are incorporated into the initial-boundary value problem (IBVP), and noise is incorporated into the mathematical
description of the phenomenon via a fractional Brownian motion process. The IBVP is subsequently reformulated as an abstract
second-order stochastic evolution equation driven by a fractional Brownian motion (fBm) dependent upon a family of probability
measures in a real separable Hilbert space and is studied using the tools of cosine function theory, stochastic analysis, and fixed-point
theory. Global existence and uniqueness results for mild solutions, continuous dependence estimates, and various approximation
results are established and applied in the context of the model.
1. Introduction
The mathematical description of the dynamic buckling has
been the subject of investigation for decades and is of interest to the engineering world. Dynamic buckling arises in
various ways including vibrations, single load pulses of large
amplitude, occurrence of a suddenly applied load, and flutter
enhanced bending (see [1]). For the purpose of our study we
restrict our attention to the dynamic buckling of a hinged
extensible beam which is stretched or compressed by an axial
force.
Dickey [2] initiated an investigation of the hyperbolic
partial differential equation
−(
𝑊 (0, 𝑡) = 𝑊 (đŋ, 𝑡) =
𝜕2 𝑊 (0, 𝑡) 𝜕2 𝑊 (đŋ, 𝑡)
=
= 0.
𝜕𝑧2
𝜕𝑧2
(2)
The initial conditions given by
𝜕2 𝑊 (𝑧, 𝑡) 𝐸đŧ 𝜕4 𝑊 (𝑧, 𝑡)
+
𝜕𝑡2
𝜌
𝜕𝑧4
đŋ
where 𝐸 is Young’s modulus, đŧ is the cross-sectional moment
of inertia, 𝜌 is the density, 𝜂 is an axial force, đŋ is the
natural length of the beam, and 𝐴 𝑐 is the cross-sectional
area; these are all positive parameters. Here, 𝑊(𝑧, 𝑡) describes
the transverse deflection of an extensible beam at point 𝑧 at
time 𝑡. The nonlinear term in (1) accounts for the change
in tension of the beam due to extensibility. The ends of
the beam are held at a fixed distance apart and the ends
are hinged; this translates to the following boundary conditions:
2
2
𝜂 𝐸𝐴 𝑐
𝜕 𝑊 (𝜉, 𝑡)
𝜕 𝑊 (𝑧, 𝑡)
𝑑𝜉)
= 0,
+
∫
𝜌 2𝜌đŋ 0
𝜕𝜉2
𝜕𝑧2
0 ≤ 𝑧 ≤ đŋ, 0 < 𝑡,
(1)
𝑊 (𝑧, 0) = 𝜍0 (𝑧) ,
𝜕𝑊 (𝑧, 0)
= 𝜍1 (𝑧) ,
𝜕𝑡
0≤𝑧≤đŋ
(3)
describe the initial deflection and initial velocity at each
point 𝑧 of the beam.
Fitzgibbon [3] and Ball [4] established the general existence theory for IBVP (1)–(3). Patcheu [5] subsequently
2
incorporated a nonlinear friction force term into the mathematical model to account for dissipation; this was done by
replacing the right side of (1) by the term 𝜁 |(𝜕𝑊(𝑧, 𝑡))/𝜕𝑡|,
where 𝜁 is a bounded linear operator. More recently,
Balachandran and Park [6] further introduced the term
− 𝜆 (𝜕4 𝑊(𝑧, 𝑡)/𝜕𝑡2 𝜕𝑧2 ) into (1) to account for the fact that
during vibration, the elements of the beam not only undergo
translator motion but also rotate.
The above results were established for the deterministic case and did not account for environmental noise. As
Kannan and Bharucha-Reid [7] points out, the variability in
measurement obtained experimentally suggests that studying
a stochastic version of the model is advantageous. Indeed,
doing so enables us to understand the effects of noise on
the behavior of the phenomenon. To this end, Mahmudov
and McKibben [8] studied a stochastic version of (1) (with
damping), assuming that noise was incorporated into the
model via a standard one-dimensional Brownian motion
process.
The purpose of the present work is to generalize those
results in two ways. One of the directions for this generalization is that we broaden the class of nonlinear forcing
terms incorporated into the model. We do so in two ways,
one of which is to consider more general functional-type
forcing terms that capture general integral terms, semilinear terms, and others under the same abstract form.
For the second way, we point out that accounting for
certain types of nonlinearities in the mathematical modeling of some phenomena—for instance, nonlinear waves
and the dynamic buckling of a hinged extensible beam—
requires that the nonlinearities depend on the probability
distribution of the solution process. This notion was first
studied in the finite-dimensional setting [9, 10], was initiated in the infinite-dimensional setting by Ahmed and
Ding [11], and subsequently was studied by various authors
[12, 13].
The second direction of the generalization is to incorporate noise into the model via a more general fractional Brownian motion process. Regarding the mathematical
description of the noise term, it is often the case that the
standard Brownian motion is insufficient. Indeed, it has been
shown that certain processes arising in a broad array of
applications—such as communication networks, self-similar
protein dynamics, certain financial models, and nonlinear
waves—exhibit a self-similarity property in the sense that the
processes {đ‘Ĩ(đ›ŧ𝑡) : 0 ≤ 𝑡 ≤ 𝑇} and {đ›ŧđģđ‘Ĩ(𝑡) : 0 ≤ 𝑡 ≤ 𝑇} have
the same law (see [14–16]). Indeed, while the case when đģ =
1/2 generates a standard Brownian motion, concrete data
from a variety of applications have exhibited other values of
đģ, and it seems that this difference enters in a nonnegligible
way in the modeling of this phenomena. In fact, since đĩđģ(𝑡)
is not a semimartingale unless đģ = 1/2, the standard
stochastic calculus involving the ItoĖ integral cannot be used in
the analysis of related stochastic evolution equations. Several
authors have studied the formulation of stochastic calculus
for fBm and differential/evolution equations driven by fBm
in the past decade [14–18].
International Journal of Stochastic Analysis
Specifically, we consider the following stochastic IBVP:
𝜕(
𝜕3 𝑊 (𝑧, 𝑡; 𝜔)
𝐸đŧ 𝜕4 𝑊 (𝑧, 𝑡; 𝜔)
𝜕𝑊 (𝑧, 𝑡; 𝜔)
+
đ›ŧ
] 𝜕𝑡
)+[
𝜕𝑡
𝜌
𝜕𝑧4
𝜕𝑡 𝜕𝑧2
− [(
đģ 𝐸𝐴 𝑐 đŋ ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨ 𝜕𝑊 (𝜉, 𝑡; 𝜔) ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨2
+
∫ ķĩ„¨
ķĩ„¨ķĩ„¨ 𝑑𝜉)
ķĩ„¨ķĩ„¨
𝜌 2𝜌đŋ 0 ķĩ„¨ķĩ„¨ķĩ„¨
𝜕𝜉
×
𝜕2 𝑊 (𝑧, 𝑡; 𝜔)
] 𝜕𝑡
𝜕𝑧2
= F (𝑊 (𝑧, 𝑡; 𝜔) , 𝑡; 𝜔) 𝜕𝑡 + 𝑔 (𝑧, 𝑡; 𝜔) 𝑑đ›Ŋđģ (𝑡; 𝜔) ,
(4)
𝑊 (0, 𝑡; 𝜔) = 𝑊 (đŋ, 𝑡; 𝜔) =
𝜕2 𝑊 (0, 𝑡; 𝜔) 𝜕2 𝑊 (đŋ, 𝑡; 𝜔)
=
= 0,
𝜕𝑧2
𝜕𝑧2
(5)
𝜕𝑊 (𝑧, 0; 𝜔)
= 𝜍1 (𝑧; 𝜔) ,
𝜕𝑡
𝑊 (𝑧, 0; 𝜔) = 𝜍0 (𝑧; 𝜔) ,
(6)
where 0 ≤ 𝑧 ≤ đŋ, 0 ≤ 𝑡 ≤ 𝑇, 𝜔 ∈ Ω (a complete probability
space), {đ›Ŋđģ(𝑡; 𝜔) : 0 ≤ 𝑡 ≤ 𝑇} is a fractional Brownian
motion, and the nonlinear forcing term F(𝑊(𝑧, 𝑡; 𝜔), 𝑡; 𝜔)
will take on various forms including
𝑡
∫ 𝑏 (𝑡, 𝑠) 𝐹0 (𝑠, 𝑊 (𝑧, 𝑠; 𝜔)) 𝑑𝑠 + 𝐹1 (𝑡, 𝑊 (𝑧, 𝑡; 𝜔)) ,
0
𝑇
𝑠
0
0
(7)
∫ 𝑎 (𝑡, 𝑠) 𝐹2 (𝑧, 𝑠, 𝑊 (𝑧, 𝑠; 𝜔) , ∫ 𝑘 (𝑠, 𝜏, 𝑊 (𝑧, 𝜏; 𝜔)) 𝑑𝜏) 𝑑𝑠,
𝐹3 (𝑧, 𝑡, 𝑊 (𝑧, 𝑡; 𝜔)) + ∫
L2 (0,đŋ)
(8)
𝐹4 (𝑧, đ‘Ļ, 𝑡; 𝜔) 𝜇 (𝑧, 𝑡) (𝑑đ‘Ļ) ,
(9)
where {𝜇(𝑧, 𝑡) : 0 ≤ 𝑧 ≤ đŋ, 0 ≤ 𝑡 ≤ 𝑇} is a family of probability
measures, and 𝑎, 𝑘, and 𝐹𝑖 (𝑖 = 1, 2, 3, 4) are appropriate
mappings.
The plan of the paper is as follows. We will collect
some preliminary information on cosine function theory,
special function spaces, probability measures, and fractional
Brownian motion in Section 2. Then we present abstract formulations of the IBVPs under consideration in Section 3. The
main theoretical results together with corollaries illustrating
the applicability to the IBVPs introduced above are gathered
in Section 4. The proofs of the main results are provided in
Section 5, followed by an interpretation of these results for
the specific nonlinear beam model in Section 6. An extensive
reference list then follows.
2. Preliminaries
For details of this section, we refer the reader to [4, 16, 17,
19, 20] and the references therein. Throughout this paper,𝑈
and 𝑉 are real separable Hilbert spaces with norms ‖ ⋅ ‖𝑈 and
‖ ⋅ ‖𝑉, and inner products ⟨⋅, ⋅⟩𝑈 and ⟨⋅, ⋅⟩𝑉 equipped with a
complete orthonormal basis {𝑒𝑗 |𝑗 = 1, 2, . . .}. Also,(Ω, F, 𝑃)
International Journal of Stochastic Analysis
3
is a complete probability space. Henceforth, for brevity,
we will suppress the dependence of random variables on
𝜔.
We make use of several different function spaces throughout this paper. BL(𝑈) is the space of all bounded linear
operators on 𝑈, while L2 (𝑈) = L2 ((0, 𝑇); 𝑈) stands for the
space of all 𝑈-valued random variables đ‘Ļ for which 𝐸‖đ‘Ļ‖2𝑈 <
∞. Also, đļ([0, 𝑇]; 𝑈) stands for the space of L2 -continuous
𝑈-valued random variables đ‘Ļ : [0, 𝑇] → 𝑈 such that
ķĩ„Š2
ķĩ„Šķĩ„Š ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Šķĩ„Šđ‘Ļķĩ„Šķĩ„Šđļ([0,𝑇];𝑈) ≡ sup 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ļ (𝑡)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ < ∞.
0≤𝑡≤𝑇
(10)
The space of Hilbert Schmidt operators from 𝑉 into 𝑈 is
denoted by L(𝑉, 𝑈) and is equipped with the norm ‖ ⋅ ‖L(𝑉,𝑈) .
The remaining function spaces coincide with those used
in [11]; we recall them here for convenience. First, B(𝑈)
stands for the Borel class on 𝑈 and ℘(𝑈) represents the space
of all probability measures defined on B(𝑈) equipped with
the weak convergence topology. Define 𝜆 : 𝑈 → (0, ∞) by
𝜆(đ‘Ĩ) = 1 + ‖đ‘Ĩ‖𝑈, đ‘Ĩ ∈ 𝑈, and consider the space
ķĩ„Š ķĩ„Š
ķĩ„¨
C (𝑈) = { 𝜑 : 𝑈 ķŗ¨€→ 𝑈ķĩ„¨ķĩ„¨ķĩ„¨ 𝜑 is continuous and ķĩ„Šķĩ„Šķĩ„Šđœ‘ķĩ„Šķĩ„Šķĩ„ŠC
ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šđœ‘ (đ‘Ĩ) − 𝜑 (đ‘Ļ)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šđœ‘ (đ‘Ĩ)ķĩ„Šķĩ„Š
= sup ķĩ„Š 2 ķĩ„Šđ‘ˆ + sup ķĩ„Š ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
đ‘Ĩ ∈ 𝑈 𝜆 (đ‘Ĩ)
đ‘Ĩ =Ė¸ đ‘Ļ in 𝑈
(11)
℘𝑠𝜆𝑝 (𝑈)
= {𝑚 : 𝑈 ķŗ¨€→ R | 𝑚 is assigned measure on 𝑈 such that
‖𝑚‖𝜆𝑝 = ∫ 𝜆𝑝 (đ‘Ĩ) |𝑚| (𝑑đ‘Ĩ) < ∞} ,
(12)
−
where 𝑚 = 𝑚 − 𝑚 is the Jordan decomposition of 𝑚 and
|𝑚| = 𝑚+ + 𝑚− . Then define the space ℘𝜆2 (𝑈) = ℘𝑠𝜆2 (𝑈) ∩
℘(𝑈) equipped with the metric 𝜌 given by
𝜌 (𝜍1 , 𝜍2 )
ķĩ„Š ķĩ„Š
= sup {∫ 𝜑 (đ‘Ĩ) (𝜍1 − 𝜍2 ) (𝑑đ‘Ĩ) : ķĩ„Šķĩ„Šķĩ„Šđœ‘ķĩ„Šķĩ„Šķĩ„ŠC ≤ 1} .
(13)
𝑈
It is known that (℘𝜆2 (𝑈), 𝜌) is a complete metric space. The
space of all continuous ℘𝜆2 (𝑈)-valued functions defined on
[0, 𝑇] is denoted by C𝜆2 = C𝜆2 ([0, 𝑇]; (℘𝜆2 (𝑈), 𝜌)) is complete when equipped with the metric
𝐷𝑇 (𝜍1 , 𝜍2 ) = sup 𝜌 (𝜍1 (𝑡) , 𝜍2 (𝑡)) ,
𝑡 ∈ [0,𝑇]
(c) đļ(𝑡 + 𝑠) + đļ(𝑡 − 𝑠) = 2đļ(𝑡)đļ(𝑠), for all 𝑡, 𝑠 ∈ R, is called
a strongly continuous cosine family.
(ii) The corresponding strongly continuous sine family
𝑡
{𝑆(𝑡) : 𝑡 ∈ R} ⊂ BL(𝑈) is defined by 𝑆(𝑡)𝑧 = ∫0 đļ(𝑠)𝑧 𝑑𝑠,
for all 𝑡 ∈ R, for all 𝑧 ∈ 𝑈.
(iii) The (infinitesimal) generator 𝐴 : 𝑈 → 𝑈 of {đļ(𝑡) :
𝑡 ∈ R} is given by 𝐴𝑧 = (𝑑2 /𝑑𝑡2 )đļ(𝑡)𝑧|𝑡 = 0 , for all 𝑧 ∈ 𝐷(𝐴) =
{𝑧 ∈ 𝑈 : đļ(⋅)𝑧 ∈ đļ2 (R; 𝑈)}.
It is known that the infinitesimal generator 𝐴 is a closed,
densely-defined operator on 𝑈 (see [21]). Such a cosine and
(corresponding sine) family and its generator satisfy the
following properties.
Proposition 2. Suppose that 𝐴 is the infinitesimal generator
of a cosine family of operators {đļ(𝑡) : 𝑡 ∈ R} (cf. Definition 1).
Then the following hold.
(i) There exist 𝑀𝐴 ≥ 1 and 𝜔 ≥ 0 such that ‖đļ(𝑡)‖ ≤
𝑀𝐴 𝑒𝜔|𝑡| and hence, ‖𝑆(𝑡)‖ ≤ 𝑀𝐴 𝑒𝜔|𝑡| ,
𝑟
For 𝑝 ≥ 1, we let
+
(b) đļ(𝑡)𝑧 is continuous in 𝑡 on R, for all 𝑧 ∈ 𝑈,
(ii) 𝐴 ∫𝑠 𝑆(đ‘ĸ)𝑧𝑑đ‘ĸ = [đļ(𝑟) − đļ(𝑠)]𝑧, for all 0 ≤ 𝑠 ≤ 𝑟 < ∞,
< ∞} .
𝑈
(a) đļ(0) = đŧ,
∀𝜍1 , 𝜍2 ∈ C𝜆2 . (14)
We recall some facts about cosine families of operators defined as follows.
Definition 1. (i) The one-parameter family {đļ(𝑡) : 𝑡 ∈ R} ⊂
BL(𝑈), satisfying
(iii) There𝑟 exists 𝑁 ≥ 1 such that ‖𝑆(𝑠) − 𝑆(𝑟)‖ ≤
𝑁| ∫𝑠 𝑒 𝑤|𝑠| 𝑑𝑠|, for all 0 ≤ 𝑠 ≤ 𝑟 < ∞.
The Uniform Boundedness Principle, together with (i)
above, implies that both {đļ(𝑡) : 𝑡 ∈ [0, 𝑇]} and{𝑆(𝑡) : 𝑡 ∈
[0, 𝑇]} are uniformly bounded by 𝑀∗ = 𝑀𝐴 𝑒𝜔𝑇 .
Next, we make precise the definition of a 𝑈-valued
fBm and related stochastic integral used in this paper. The
approach we use coincides with the one formulated and
∞
analyzed in [15, 17]. Let {đ›Ŋ𝑗đģ(𝑡)| 𝑡 ≥ 0} be a sequence of
𝑗=1
independent, one-dimensional fBms with Hurst parameter
đģ ∈ (1/2, 1) such that for all 𝑗 = 1, 2, . . .
(i) đ›Ŋ𝑗đģ(0) = 0,
2
(ii) 𝐸[đ›Ŋ𝑗đģ(𝑡) − đ›Ŋ𝑗đģ(𝑠)] = |𝑡 − 𝑠|2đģ]𝑗 ,
2
(iii) 𝐸[đ›Ŋ𝑗đģ(1)] = ]𝑗 > 0,
(iv) ∑∞
𝑗=1 ]𝑗 < ∞.
2
đģ
2đģ ∞
In such case, ∑∞
∑𝑗=1 ]𝑗 < ∞, so
𝑗=1 𝐸‖đ›Ŋ𝑗 (𝑡)𝑒𝑗 ‖𝑈 = 𝑡
that the following definition is meaningful.
đģ
Definition 3. For every 𝑡 ≥ 0, đ›Ŋđģ(𝑡) = ∑∞
𝑗=1 đ›Ŋ𝑗 (𝑡)𝑒𝑗 is a 𝑈valued fBm, where the convergence is understood to be in the
mean-square sense.
It has been shown in [17] that the covariance operator of
{đ›Ŋđģ(𝑡) : 𝑡 ≥ 0} is a positive nuclear operator 𝑄 such that
tr 𝑄 (𝑡, 𝑠) =
1∞
∑] [𝑡2đģ + 𝑠2đģ − |𝑡 − 𝑠|2đģ] .
2 𝑗=1 𝑗
(15)
We now outline the discussion leading to the definition of
the stochastic integral associated with {đ›Ŋđģ(𝑡) : 𝑡 ≥ 0} for
4
International Journal of Stochastic Analysis
bounded, measurable functions. To begin, assume that 𝑔 :
[0, 𝑇] → L(𝑉, 𝑈) is a simple function; that is, there exists
{𝑔𝑖 : 𝑖 = 1, . . . , 𝑛} ⊆ R such that
𝑔 (𝑡) = 𝑔𝑖 ,
∀ 𝑡𝑖−1 ≤ 𝑡 ≤ 𝑡𝑖 ,
(16)
where 0 = 𝑡0 < 𝑡1 < ⋅ ⋅ ⋅ < 𝑡𝑛−1 < 𝑡𝑛 = 𝑇 and max1≤𝑖≤𝑛
‖𝑔𝑖 ‖L(𝑉,𝑈) = 𝐾.
𝑇
Definition 4. The 𝑈-valued stochastic integral ∫0 𝑔(𝑡)𝑑đ›Ŋđģ(𝑡)
is defined by
𝑇
∞
𝑇
0
𝑗=1
0
∞
𝑛
∫ 𝑔 (𝑡) 𝑑đ›Ŋđģ (𝑡) = ∑ (∫ 𝑔 (𝑡) 𝑑đ›Ŋ𝑗đģ (𝑡)) 𝑒𝑗
=
∑ (∑𝑔𝑖 [đ›Ŋ𝑗đģ (𝑡𝑖 )
𝑗=1 𝑖=1
(17)
−
đ›Ŋ𝑗đģ (𝑡𝑖−1 )]) 𝑒𝑗 .
As argued in Lemma 2.2 of [17], this integral is well-defined
since
ķĩ„Šķĩ„Š 𝑇
ķĩ„Šķĩ„Š2
∞
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫ 𝑔 (𝑡) 𝑑đ›Ŋđģ (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ≤ 𝐾2 𝑇2đģ ∑]𝑗 < ∞.
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
𝑗=1
>
0 for which
(i) there exists 𝑀𝐹4
‖𝐹4 (đ‘Ļ, 𝑡, 𝑤)‖L2 (0,đŋ) ≤ 𝑀𝐹4 (1 + ‖𝑤‖L2 (0,đŋ) ),
for all đ‘Ļ ∈ (0, đŋ), 𝑡 ∈ [0, 𝑇], and 𝑤 ∈ L2 (0, đŋ);
(ii) 𝐹4 (đ‘Ļ, 𝑡, ⋅) : L2 (0, đŋ) → L2 (0, đŋ) belongs to
C𝜆2 , for every đ‘Ļ ∈ (0, đŋ), 𝑡 ∈ [0, 𝑇];
(A3) 𝑔 : (0, đŋ) × [0, 𝑇] → L(L2 (0, đŋ), L2 (0, đŋ)) is a bounded measurable function;
(A4) 𝜇(⋅, 𝑡) ∈ ℘𝜆2 (L2 (0, đŋ)) is the probability law of 𝑊(⋅, 𝑡),
for each 𝑡 ∈ [0, 𝑇];
(A5) 𝜍0 (⋅), 𝜍1 (⋅) ∈ L2 (Ω; L2 (0, đŋ)).
We reformulate the associated IBVP by making the following identifications as in [8]; we recall the highlights of that
formulation here for completeness of the discussion.
Let 𝑈 = 𝑉 = L2 (0, đŋ) (note they are real separable Hilbert spaces) and identify the solution process đ‘Ĩ : [0, 𝑇] → 𝑈
by
đ‘Ĩ (𝑡) (𝑧) = 𝑊 (𝑧, 𝑡) ,
∀𝑧 ∈ (0, đŋ) , 𝑡 ∈ [0, 𝑇] .
(19)
Define the operator 𝐴 : 𝐷(𝐴) ⊂ 𝑈 → 𝑈 by
(18)
Since the set of simple functions is dense in the space of
bounded, measurable L(𝑉, 𝑈)-valued functions, a standard
density argument can be used to extend Definition 4 to the
case of a general bounded measurable integrand.
Finally, in addition to the familiar Young, HoĖˆlder,
and Minkowski inequalities, the inequality of the form
𝑚
(∑𝑛𝑖 = 1 𝑎𝑖 ) ≤ 𝑛𝑚−1 ∑𝑛𝑖 = 1 𝑎𝑖𝑚 , where 𝑎𝑖 is a nonnegative constant (𝑖 = 1, . . . , 𝑛) and 𝑚, 𝑛 ∈ N, will be used to establish
various estimates.
3. Abstract Formulation
The goal of this section is to reformulate the stochastic IBVPs
introduced in Section 1 as abstract stochastic evolution equations. There will be two related abstract formulations
depending on the nature of the nonlinearity. We begin with
the reformulation (4)–(6) assuming that the forcing term
F(𝑊(𝑧, 𝑡; 𝜔), 𝑡; 𝜔) is given by (9).
We impose the following conditions:
(A1) 𝐹3 : (0, đŋ)×[0, 𝑇]×R → R satisfies the Caratheodory
conditions (i.e., measurable in (đ‘Ĩ, 𝑡) and continuous
in the third variable) such that
(i) there exists 𝑀𝐹3 > 0 for which |𝐹3 (đ‘Ļ, 𝑡, 𝑤)| ≤
𝑀𝐹3 (1 + |𝑤|), for all đ‘Ļ ∈ (0, đŋ), 𝑡 ∈ [0, 𝑇], and
𝑤 ∈ R;
(ii) there exists 𝑀𝐹3 > 0 for which |𝐹3 (đ‘Ļ, 𝑡, 𝑤1 ) −
𝐹3 (đ‘Ļ, 𝑡, 𝑤2 )| ≤ 𝑀𝐹3 |𝑤1 − 𝑤2 |, for all đ‘Ļ ∈ (0, đŋ),
𝑡 ∈ [0, 𝑇], and 𝑤1 , 𝑤2 ∈ R;
(A2) 𝐹4 : (0, đŋ) × [0, 𝑇] × L2 (0, đŋ) → L2 (0, đŋ) satisfies the
Caratheodory conditions such that
𝐴đ‘Ĩ (𝑡, ⋅) =
𝜕4 đ‘Ĩ (𝑡, ⋅)
,
𝜕𝑧4
𝐷 (𝐴) = {đ‘Ĩ ∈ đģ4 (0, đŋ) : đ‘Ĩ (0) = đ‘Ĩ (đŋ) = đ‘Ĩķ¸€ ķ¸€  (0) = đ‘Ĩķ¸€ ķ¸€  (đŋ) = 0} .
(20)
It is known that 𝐴 is a positive, self-adjoint operator on 𝑈 (cf.
[21, 22]). As such, 𝐴 generates a strongly continuous cosine
family of operators on 𝑈. We will express the other terms on
the left side of (4) using fractional powers of 𝐴, as in [22]. To
this end, the eigenvalues of 𝐴 are {𝜆 𝑛 = (𝑛𝜋)4 : 𝑛 = 1, 2, 3, . . .}
with corresponding eigenvalues
{𝑧𝑛 (𝑠) =
√2
sin (𝑛𝜋𝑠) : 𝑛 = 1, 2, 3, . . . , 𝑠 ∈ [0, đŋ]} .
đŋ
(21)
Then 𝐴 has spectral representation
∞
𝐴đ‘Ļ = ∑ 𝜆 𝑛 ⟨đ‘Ļ, 𝑧𝑛 ⟩ 𝑧𝑛 .
(22)
𝑛=1
The following operators are well defined because the fractional powers of 𝐴 are positive and self-adjoint:
∞
𝐴1/2 đ‘Ļ = ∑ 𝜆1/2
𝑛 ⟨đ‘Ļ, 𝑧𝑛 ⟩ 𝑧𝑛 = −
𝑛=1
1/4
𝐴
đ‘Ļ=
∞
∑ 𝜆1/4
𝑛
𝑛=1
𝜕2 đ‘Ļ
,
𝜕𝑧2
𝜕đ‘Ļ
⟨đ‘Ļ, 𝑧𝑛 ⟩ 𝑧𝑛 =
.
𝜕𝑧
(23)
Also, observe that
ķĩ„Šķĩ„Š 1/4 ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Šđ´ đ‘Ļķĩ„Šķĩ„Š = ⟨𝐴1/4 đ‘Ļ, 𝐴1/4 đ‘Ļ⟩
ķĩ„Š
ķĩ„Šđ‘ˆ
∞
ķĩ„Šķĩ„Š2
đŋķĩ„Š
ķĩ„Šķĩ„Š 𝜕đ‘Ļ
2
ķĩ„Š (𝑤, 𝑡)ķĩ„Šķĩ„Šķĩ„Š 𝑑𝑤.
= ∑ 𝜆1/2
𝑛 ⟨đ‘Ļ, 𝑧𝑛 ⟩ = ∫ ķĩ„Š
ķĩ„Š 𝜕𝑧
ķĩ„Šķĩ„Š
0 ķĩ„Š
ķĩ„Š
ķĩ„Šđ‘ˆ
𝑛=1
(24)
International Journal of Stochastic Analysis
5
It can be shown using standard computations involving the
inner product and properties of fractional powers of 𝐴
2
2
that ‖𝐴1/4 đ‘Ĩ(𝑡)‖𝑈 and ‖𝐴1/2 đ‘Ĩ(𝑡)‖𝑈 are uniformly bounded on
[0, 𝑇].
Define the mappings 𝑓 : [0, 𝑇] × đ‘ˆ × ℘𝜆2 (𝑈) → 𝑈, 𝑔 :
[0, 𝑇] → L(𝑉, 𝑈), đ‘Ĩ0 (⋅), and đ‘Ĩ1 (⋅), and the operator đĩ : 𝑈 →
𝑈 as follows:
will describe such nonlinear forcing terms more generally as
a so-called functional. Precisely, we consider the following
abstract second-order functional stochastic evolution equation in a real separable Hilbert space 𝑈:
𝑑đ‘Ĩķ¸€  (𝑡) = (đĩđ‘Ĩķ¸€  (𝑡) + 𝐴đ‘Ĩ (𝑡) + F (đ‘Ĩ) (𝑡)) 𝑑𝑡
+ 𝑔 (𝑡) 𝑑đ›Ŋđģ (𝑡) ,
𝑓 (𝑡, đ‘Ĩ (𝑡) , 𝜇 (𝑡)) (𝑧)
đ‘Ĩ (0) = đ‘Ĩ0 ,
= 𝐹3 (𝑧, 𝑡, 𝑊 (𝑧, 𝑡))
+ ∫
L2 (0,đŋ)
−(
đģ 𝐸𝐴 𝑐 đŋ ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨ 𝜕𝑊 (𝜉, 𝑡) ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨2
𝜕2 𝑊 (𝑧, 𝑡)
+
∫ ķĩ„¨ķĩ„¨
ķĩ„¨ķĩ„¨ 𝑑𝜉)
𝜌 2𝜌đŋ 0 ķĩ„¨ķĩ„¨ 𝜕𝜉 ķĩ„¨ķĩ„¨
𝜕𝑧2
L2 (0,đŋ)
−(
𝐹4 (𝑧, đ‘Ļ, 𝑡) 𝜇 (𝑧, 𝑡) (𝑑đ‘Ļ)
đģ 𝐸𝐴 𝑐 ķĩ„Šķĩ„Š 1/4
ķĩ„Š2
ķĩ„Šķĩ„Šđ´ đ‘Ĩ (𝑡)ķĩ„Šķĩ„Šķĩ„Š ) 𝐴1/2 đ‘Ĩ (𝑡) ,
+
ķĩ„Šđ‘ˆ
ķĩ„Š
𝜌 2𝜌đŋ
𝜕𝑊 (𝑧, 𝑡)
𝜕3 𝑊 (𝑧, 𝑡)
= đ›ŧ𝐴1/2 (
),
2
𝜕𝑡𝜕𝑧
𝜕𝑡2
đ‘Ĩ0 (0) (𝑧) = 𝜍0 (𝑧) ,
(25)
for all 𝑧 ∈ (0, đŋ) and 𝑡 ∈ [0, 𝑇].
These identifications can be used to formulate (4)–(6),
with nonlinear forcing term given by (9) as the following
abstract second-order stochastic evolution equation in a real
separable Hilbert space 𝑈:
𝑑đ‘Ĩķ¸€  (𝑡) = (đĩđ‘Ĩķ¸€  (𝑡) + 𝐴đ‘Ĩ (𝑡) + 𝑓 (𝑡, đ‘Ĩ (𝑡) , 𝜇 (𝑡))) 𝑑𝑡
đ‘Ĩ (0) = đ‘Ĩ0 ,
0 ≤ 𝑡 ≤ 𝑇,
đ‘Ĩķ¸€  (0) = đ‘Ĩ1 ,
(A6) {𝑏(𝑡, 𝑠) : 0 ≤ 𝑡 ≤ 𝑠 ≤ 𝑇} ⊂ BL(𝑈);
(A7) 𝐹𝑖 : (0, đŋ) × [0, 𝑇] × R → R (𝑖 = 0, 1) are mappings
that satisfy the Caratheodory conditions and are such
that there exists 𝑀𝐹𝑖 > 0 such that
ķĩ„¨ķĩ„¨ķĩ„¨đšđ‘– (đ‘Ļ, 𝑡, 𝑤1 ) − 𝐹𝑖 (đ‘Ļ, 𝑡, 𝑤2 )ķĩ„¨ķĩ„¨ķĩ„¨ ≤ 𝑀𝐹 ķĩ„¨ķĩ„¨ķĩ„¨đ‘¤1 − 𝑤2 ķĩ„¨ķĩ„¨ķĩ„¨ ,
(28)
ķĩ„¨
ķĩ„¨
ķĩ„¨
𝑖 ķĩ„¨
for all đ‘Ļ ∈ (0, đŋ), 𝑡 ∈ [0, 𝑇], and 𝑤1 , 𝑤2 , 𝑧1 , 𝑧2 ∈ R;
đ‘Ĩ1 (0) (𝑧) = 𝜍1 (𝑧) ,
+ 𝑔 (𝑡) 𝑑đ›Ŋđģ (𝑡) ,
đ‘Ĩķ¸€  (0) = đ‘Ĩ1 ,
for all đ‘Ļ ∈ (0, đŋ), 𝑡 ∈ [0, 𝑇], and 𝑤1 , 𝑤2 ∈ R;
(A8) 𝐹2 : (0, đŋ) × [0, 𝑇] × R × R → R satisfies the Caratheodory conditions and is such that there exists
𝑀𝐹2 > 0 such that
ķĩ„¨ķĩ„¨
ķĩ„¨
ķĩ„¨ķĩ„¨đš2 (đ‘Ļ, 𝑡, 𝑤1 , 𝑧1 ) − 𝐹2 (đ‘Ļ, 𝑡, 𝑤2 , 𝑧2 )ķĩ„¨ķĩ„¨ķĩ„¨
(29)
ķĩ„¨ ķĩ„¨
ķĩ„¨
ķĩ„¨
≤ 𝑀𝐹2 [ķĩ„¨ķĩ„¨ķĩ„¨đ‘¤1 − 𝑤2 ķĩ„¨ķĩ„¨ķĩ„¨ + ķĩ„¨ķĩ„¨ķĩ„¨đ‘§1 − 𝑧2 ķĩ„¨ķĩ„¨ķĩ„¨] ,
𝑔 (𝑡) (𝑧) = 𝑔 (𝑧, 𝑡) ,
đĩ (đ‘Ĩķ¸€  (𝑡)) (𝑧) = đ›ŧ
(27)
where F : đļ([0, 𝑇]; 𝑈) → L2 (0, 𝑇; L2 (Ω; 𝑈)) is a given
functional and all other identifications are the same as for
(26). We impose the following conditions for (7) and (8):
𝐹4 (𝑧, đ‘Ļ, 𝑡) 𝜇 (𝑧, 𝑡) (𝑑đ‘Ļ)
= 𝐹3 (𝑧, 𝑡, 𝑊 (𝑧, 𝑡)) + ∫
0 ≤ 𝑡 ≤ 𝑇,
(A9) 𝑎 ∈ L2 ((0, 𝑇)2 ; R);
(A10) 𝑘 : Δ × R → R, where Δ = {(𝑠, 𝑡) : 0 < 𝑠 < 𝑡 < 𝑇}
satisfies
ķĩ„¨
ķĩ„¨
ķĩ„¨
ķĩ„¨ķĩ„¨
(30)
ķĩ„¨ķĩ„¨đ‘˜ (𝑡, 𝑠, đ‘Ĩ1 ) − 𝑘 (𝑡, 𝑠, đ‘Ĩ2 )ķĩ„¨ķĩ„¨ķĩ„¨ ≤ 𝑀𝑘 ķĩ„¨ķĩ„¨ķĩ„¨đ‘Ĩ1 − đ‘Ĩ2 ķĩ„¨ķĩ„¨ķĩ„¨ ,
for all đ‘Ĩ1 , đ‘Ĩ2 ∈ R and (𝑠, 𝑡) ∈ Δ. Identify the solution
process đ‘Ĩ : [0, 𝑇] → 𝑈 by
đ‘Ĩ (𝑡) (𝑧) = 𝑊 (𝑧, 𝑡) ,
(26)
𝜇 (𝑡) = probability distribution of đ‘Ĩ (𝑡)
in a real separable Hilbert space 𝑈. (By the probability
distribution of đ‘Ĩ(𝑡), we mean 𝜇(𝑡)(𝑧) = 𝑃({𝜔 ∈ Ω : đ‘Ĩ(𝑡, 𝜔) ∈
𝑧}), for each 𝑧 ∈ B(𝑈).) Here, 𝐴 : 𝐷(𝐴) ⊂ 𝑈 → 𝑈 is a linear
(possibly unbounded) operator which generates a strongly
continuous cosine family {đļ(𝑡) : 𝑡 ≥ 0} on 𝑈 with associated
sine family {𝑆(𝑡) : 𝑡 ≥ 0}; 𝑓 : [0, 𝑇] × đ‘ˆ × ℘𝜆2 (𝑈) → 𝑈;
𝑔 : [0, 𝑇] → L(𝑉, 𝑈) is a bounded, measurable mapping;
đĩ : 𝑈 → 𝑈 is a bounded linear operator; {đ›Ŋđģ(𝑡) : 𝑡 ≥
0} is a 𝑉-valued fBm with Hurst parameter đģ ∈ (1/2, 1);
andđ‘Ĩ0 , đ‘Ĩ1 ∈ L2 (Ω; 𝑈).
We next consider the same IBVP, with the exception
that the forcing term is given by a mapping of the form (7)
or (8). While all other identifications remain identical, we
∀ 𝑧 ∈ (0, đŋ) , 𝑡 ∈ [0, 𝑇] .
(31)
It can be shown that under these assumptions, the following are well-defined mappings from đļ([0, 𝑇]; 𝑈) into L2 (0,
𝑇; L2 (Ω; 𝑈)):
𝑡
F (đ‘Ĩ) (𝑡) = ∫ 𝑎 (𝑡, 𝑠)
0
𝑠
× đš2 (⋅, 𝑠, 𝑊 (⋅, 𝑠; 𝜔) , ∫ 𝑘 (𝑠, 𝜏, 𝑊 (⋅, 𝜏; 𝜔)) 𝑑𝜏) 𝑑𝑠,
0
(32)
F (đ‘Ĩ) (𝑡)
𝑡
= ∫ 𝑏 (𝑡, 𝑠) 𝐹0 (⋅, 𝑠, 𝑊 (⋅, 𝑠; 𝜔)) 𝑑𝑠 + 𝐹1 (⋅, 𝑡, 𝑊 (⋅, 𝑡; 𝜔)) .
0
(33)
The strategy is to formulate a general theory for these two
abstract second-order problems and then to apply the results
to the specific IBVPs formulated in Section 1.
6
International Journal of Stochastic Analysis
4. Statement of Results
We consider mild solutions of (26) in the following sense.
Lemma 6. Assume that 𝑔 : [0, 𝑇] → L(𝑉, 𝑈) satisfies (A13).
Then, for all 0 ≤ 𝑡 ≤ 𝑇,
∞
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑡
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑑đĩđģ (𝑠)ķĩ„Šķĩ„Šķĩ„Š ≤ đļ𝑡 ∑ ]𝑗 ,
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 0
𝑗=1
Definition 5. A stochastic process đ‘Ĩ ∈ đļ([0, 𝑇]; 𝑈) is a mild
solution of (26) if
(i)
where đļ𝑡 is a positive constant depending on 𝑡, 𝑀𝑆 , and the
growth bound on g, and {]𝑗 : 𝑗 ∈ N} is defined as in the discussion leading to Definition 4.
đ‘Ĩ (𝑡) = 𝑆 (𝑡) đ‘Ĩ1 + (đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0
𝑡
+ ∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠
0
𝑡
+ ∫ 𝑆 (𝑡 − 𝑠) 𝑓 (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠)) 𝑑𝑠
(34)
𝑡
0
Let 𝜇 ∈ C𝜆2 be fixed and define the solution map Φ : đļ([0,
𝑇]; 𝑈) → đļ([0, 𝑇]; 𝑈) by
Φ (đ‘Ĩ) (𝑡) = 𝑆 (𝑡) đ‘Ĩ1 + (đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0
0
+ ∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠) ,
(38)
𝑡
0 ≤ 𝑡 ≤ 𝑇,
+ ∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠
0
𝑡
(ii) 𝜇(𝑡) is the probability distribution of đ‘Ĩ(𝑡), for all 0 ≤
𝑡 ≤ 𝑇.
+ ∫ 𝑆 (𝑡 − 𝑠) 𝑓 (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠)) 𝑑𝑠
0
𝑡
The following conditions on (26) are imposed on the data
and mappings in (26):
+ ∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠) ,
0
(A11) 𝐴 : 𝐷(𝐴) ⊂ 𝑈 → 𝑈 is the infinitesimal generator of
a strongly continuous cosine family {đļ(𝑡) : 𝑡 ≥ 0} on
𝑈 with associated sine family {𝑆(𝑡) : 𝑡 ≥ 0} on 𝑈;
(A12) 𝑓 : [0, 𝑇] × đ‘ˆ × ℘𝜆2 (𝑈) → 𝑈 satisfies the following.
(i) There exists a positive constant 𝑀𝑓 such that
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘“ (𝑡, 𝑧1 , 𝜍1 ) − 𝑓 (𝑡, 𝑧2 , 𝜍2 )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Š2
ķĩ„Š
≤ 𝑀𝑓 [𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘§1 − 𝑧2 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝜌2 (𝜍1 , 𝜍2 )] ,
(35)
globally on [0, 𝑇] × đ‘ˆ × ℘𝜆2 (𝑈),
(ii) there exists a positive constant 𝑀𝑓 such that
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘“ (𝑡, 𝑧, 𝜍)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝑀𝑓 [1 + 𝐸‖𝑧‖2𝑈 + ‖𝜍‖2𝜆2 ] ;
(36)
globally on [0, 𝑇] × đ‘ˆ × ℘𝜆2 (𝑈);
(A13) 𝑔 : [0, 𝑇] → L(𝑉, 𝑈) is a bounded, measurable mapping;
(A14) đĩ : 𝑈 → 𝑈 is a bounded linear operator;
(A15) {đ›Ŋđģ(𝑡) : 0 ≤ 𝑡 ≤ 𝑇} is a 𝑈-valued fBm;
(A16) đ‘Ĩ0 , đ‘Ĩ1 ∈ L2 (Ω; 𝑈) are (F0 , B(𝑈))-measurable,
where {F𝑡 : 0 ≤ 𝑡 ≤ 𝑇} is the family of 𝜎-algebras
F𝑡 generated by {đ›Ŋđģ(𝑠) : 0 ≤ 𝑠 ≤ 𝑡}.
Henceforth, we write
𝑀∗ = max ({‖𝑆 (𝑡)‖BL(𝑈) : 0 ≤ 𝑡 ≤ 𝑇}
∪ {‖đļ (𝑡)‖BL(𝑈) : 0 ≤ 𝑡 ≤ 𝑇}) ,
(37)
which is finite by (A11).
The following lemma, introduced in (me and dave) and
stated here without proof, is critical in establishing several
estimates.
5
= ∑đŧ𝑖đ‘Ĩ (𝑡) ,
(39)
0 ≤ 𝑡 ≤ 𝑇,
0 ≤ 𝑡 ≤ 𝑇.
𝑖=1
The first two integrals on the right side of (39) are taken in
the Bochner sense, while the third is defined in Section 2. The
operator Φ satisfies the following properties.
Lemma 7. If (A11)–(A16) hold, then Φ is a well-defined L2 continuous mapping.
Our first main result is as follows.
Theorem 8. If (A11)–(A16) hold, then (26) has a unique mild
solution đ‘Ĩ on [0, 𝑇] with corresponding probability law 𝜇 ∈
C𝜆2 . Further, for every 𝑝 ≥ 1, there exists a positive constant
đļ𝑝 (depending on 𝑝, 𝑇, and ]𝑗 ) such that
2𝑝
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
sup 𝐸‖đ‘Ĩ (𝑡)‖𝑈 ≤ đļ𝑝 (1 + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ) .
0≤𝑡≤𝑇
(40)
Mild solutions of (26) depend continuously on the initial
data and probability distribution of the state process in the
following sense.
Proposition 9. Assume that (A11)–(A16) hold, and let đ‘Ĩ and
đ‘Ļ be the mild solutions of (26) (as guaranteed to exist by
Theorem 8) corresponding to initial data đ‘Ĩ0 , đ‘Ĩ1 and đ‘Ļ0 , đ‘Ļ1 with
respective probability distributions 𝜇đ‘Ĩ and 𝜇đ‘Ļ . Then there exists
a positive constant 𝑀ķ¸€  such that
ķĩ„Š2
ķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ (𝑡) − đ‘Ļ (𝑡)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝑀ķ¸€  [ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 − đ‘Ļ0 ķĩ„Šķĩ„Šķĩ„ŠL2 (Ω; 𝑈)
ķĩ„Š2
ķĩ„Š
+ ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 − đ‘Ļ1 ķĩ„Šķĩ„Šķĩ„ŠL2 (Ω; 𝑈)
+ 𝐷𝑇2 (𝜇đ‘Ĩ , 𝜇đ‘Ļ )] .
(41)
International Journal of Stochastic Analysis
7
Next, we formulate a result in which a deterministic
initial-value problem is approximated by a sequence of stochastic equations of a particular form of (26) arising frequently in applications. Specifically, consider the deterministic initial-value problem
ķ¸€ ķ¸€ 
ķ¸€ 
𝑧 (𝑡) = 𝐴𝑧 (𝑡) + đĩ𝑧 (𝑡) + 𝐹 (𝑡, 𝑧 (𝑡)) ,
0 ≤ 𝑡 ≤ 𝑇,
(42)
where 𝑧0 ∈ 𝐷(𝐴), 𝑧1 ∈ 𝐸 = {đ‘ĸ ∈ 𝑈 : đļ(⋅)đ‘ĸ ∈ đļ1 (R; 𝑈)},
đĩ : 𝑈 → 𝑈 is a bounded linear operator, 𝐴 satisfies (A11),
and 𝐹 : [0, 𝑇] × đ‘ˆ → 𝑈 satisfies the following conditions:
(A17)
(i) there exists a positive constant 𝑀𝐹 such that
‖𝐹(𝑡, 𝑧1 ) − 𝐹(𝑡, 𝑧2 )‖𝑈 ≤ 𝑀𝐹 ‖𝑧1 − 𝑧2 ‖𝑈 globally on
[0, 𝑇] × đ‘ˆ;
(ii) there exists a positive constant 𝑀𝐹 such that
‖𝐹(𝑡, 𝑧)‖𝑈 ≤ 𝑀𝐹 ‖𝑧‖𝑈 globally on [0, 𝑇] × đ‘ˆ.
It is known that under these conditions, (42) has a unique
mild solution 𝑧 given by the representation formula
𝑧 (𝑡) = 𝑆 (𝑡) đ‘Ĩ1 + (đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0
𝑡
+ ∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠
(43)
0
𝑡
+ ∫ 𝑆 (𝑡 − 𝑠) 𝐹 (𝑠, đ‘Ĩ (𝑠)) 𝑑𝑠,
0
0 ≤ 𝑡 ≤ 𝑇.
𝑑đ‘Ĩ𝜀 (𝑡) = (𝐴 𝜀 đ‘Ĩ𝜀 (𝑡) + đĩđ‘Ĩ𝜀ķ¸€  (𝑡)
(A19) đĩ𝜀 : 𝑈 → 𝑈 is a bounded linear operator such that
đĩ𝜀 → đĩ in BL(𝑈) and ‖đĩ𝜀 ‖BL(𝑈) ≤ ‖đĩ‖BL(𝑈) , for all
𝜀 > 0.
(A20) 𝐹2𝜀 : [0, 𝑇] × đ‘ˆ → 𝑈 is Lipschitz in the second variable (with the same Lipschitz constant 𝑀𝐹 used for 𝐹
in (A17)), and 𝐹2𝜀 (𝑡, đ‘ĸ) → 𝐹(𝑡, đ‘ĸ) as 𝜀 → 0+ , for all
đ‘ĸ ∈ 𝑈, uniformly in 𝑡 ∈ [0, 𝑇].
(A21) 𝐹1𝜀 : [0, 𝑇] × đ‘ˆ → 𝑈 is a continuous mapping such
that ∫𝑈 𝐹1𝜀 (𝑡, 𝑧)𝜇𝜀 (𝑡)(𝑑𝑧) → 0 uniformly in 𝑡 as
𝜀 → 0+ .
(A22) 𝑔𝜀 : [0, 𝑇] → L(𝑉, 𝑈) is a bounded, measurable function such that 𝑔𝜀 (𝑡) → 0 as 𝜀 → 0+ , uniformly in
𝑡 ∈ [0, 𝑇].
Under these assumptions, the following result holds.
Theorem 10. Let 𝑧 and đ‘Ĩ𝜀 be the mild solutions of (42) and
(44) on [0, 𝑇], respectively. Then there exist a positive constant
𝜍 and a positive function 𝜓(𝜀) (which decreases to 0 as 𝜀 → 0+ )
such that 𝐸‖đ‘Ĩ𝜀 (𝑡) − 𝑧(𝑡)‖2𝑈 ≤ 𝜓(𝜀) exp(𝜍𝑡), for all 𝑡 ∈ [0, 𝑇].
Definition 11. A stochastic process đ‘Ĩ ∈ đļ([0, 𝑇]; 𝑈) is a mild
solution of (27) if
đ‘Ĩ (𝑡) = 𝑆 (𝑡) đ‘Ĩ1 + (đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0
+ ∫ 𝐹1𝜀 (𝑡, 𝑧) 𝜇𝜀 (𝑡) (𝑑𝑧) + 𝐹2𝜀 (𝑡, đ‘Ĩ𝜀 (𝑡))) 𝑑𝑡
𝑈
đ‘Ĩ𝜀 (0) = 𝑧0 ,
(45)
We now turn our attention to (27). By a mild solution of
(27), we mean the following.
For each 𝜀 > 0, consider the stochastic initial-value problem
+ 𝑔𝜀 (𝑡) 𝑑đ›Ŋđģ (𝑡) ,
ķĩ„Š
ķĩ„Š
ķĩ„Š
ķĩ„Š
max {ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡)ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) + ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡)ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) : 0 ≤ 𝑡 ≤ 𝑇} ≤ 𝑀∗ ,
for all 𝜀 > 0 and 𝑀∗ is the same bound used for the
cosine and sine families generated by 𝐴.
𝑧ķ¸€  (0) = 𝑧1 ,
𝑧 (0) = 𝑧0 ,
{𝑆𝜀 (𝑡) : 𝑡 ≥ 0} satisfying 𝑆𝜀 (𝑡) → 𝑆(𝑡) strongly as
𝜀 → 0+ , uniformly in 𝑡 ∈ [0, 𝑇]. Moreover,
0 ≤ 𝑡 ≤ 𝑇,
𝑡
𝑡
0
0
+ ∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠 + ∫ 𝑆 (𝑡 − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠
𝑡
+ ∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠) ,
0
đ‘Ĩ𝜀ķ¸€  (0) = 𝑧1 ,
𝜇𝜀 (𝑡) = probability distribution of đ‘Ĩ𝜀 (𝑡) ,
(44)
in 𝑈. Here, we assume that 𝑧0 ∈ 𝐷(𝐴 𝜀 ) = 𝐷(𝐴) and 𝑧1 ∈ 𝐸𝜀 =
𝐸, for all 𝜀 > 0 and that 𝐹𝑖𝜀 : [0, 𝑇] × đ‘ˆ → 𝑈 (𝑖 = 1, 2) are
given mappings. We impose the following conditions on the
data and mappings in (44), for each 𝜀 > 0.
(A18) 𝐴 𝜀 : 𝐷(𝐴 𝜀 ) = 𝐷(𝐴) → 𝐷(𝐴 𝜀 ) generates a strongly
continuous cosine family {đļ𝜀 (𝑡) : 𝑡 ≥ 0} satisfying
đļ𝜀 (𝑡) → đļ(𝑡) strongly as 𝜀 → 0+ , uniformly
in 𝑡 ∈ [0, 𝑇]. Also, the corresponding sine family
(46)
for all 0 ≤ 𝑡 ≤ 𝑇.
We assume that conditions (A11), (A13)–(A16) hold, in
addition to the following condition on the functional forcing
term:
(A23) F : đļ([0, 𝑇]; 𝑈) → L2 (0, 𝑇; L2 (Ω; 𝑈)) is such that
there exists 𝑀F > 0 for which
ķĩ„Š
ķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„ŠF (đ‘Ĩ) − F (đ‘Ļ)ķĩ„Šķĩ„Šķĩ„ŠL2 ≤ 𝑀F ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„ŠC ,
∀ đ‘Ĩ, đ‘Ļ ∈ đļ ([0, 𝑇] ; 𝑈) .
(47)
8
International Journal of Stochastic Analysis
Define the solution map Γ : đļ([0, 𝑇]; 𝑈) → đļ([0, 𝑇]; 𝑈) by
Γ (đ‘Ĩ) (𝑡) = 𝑆 (𝑡) đ‘Ĩ1 + (đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0
𝑡
+ ∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠
0
𝑡
+ ∫ 𝑆 (𝑡 − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠
0
𝑡
(48)
đģ
+ ∫ 𝑔 (𝑠) 𝑑đ›Ŋ (𝑠) ,
0 ≤ 𝑡 ≤ 𝑇,
0
+ ∫ 𝑆 (𝑡 − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠,
5
𝑖=1
(50)
0 ≤ 𝑡 ≤ 𝑇.
0
The strong continuity of đļ(𝑡) and 𝑆(𝑡) implies that
(cf. (39)). The operator Γ satisfies the following properties.
Lemma 12. If (A11), (A13)–(A16), and (A23) hold, then Γ is a
well-defined L2 -continuous mapping.
The proof of the following theorem follows almost immediately from this result.
Theorem 13. If (A11), (A13)–(A16), and (A23) hold, then
(27) has a unique mild solution đ‘Ĩ on [0, 𝑇] provided that
2 + 𝑀2 ) < 1. Further, for every 𝑝 ≥ 1, there exists
𝑀∗ √2𝑇(𝑀F
đĩ
a positive constant đļ𝑝 (depending on 𝑝, 𝑇, and ]𝑗 ) such that
2𝑝
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
sup 𝐸‖đ‘Ĩ (𝑡)‖𝑈 ≤ đļ𝑝 (1 + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ) .
0≤𝑡≤𝑇
(49a)
1/2
𝑇1/2 < 1.
(49b)
1/2
] + 𝑀đĩ2 ]
for each 0 ≤ 𝑡 ≤ 𝑇.
Using (A11) and (A14), together with HoĖˆlder’s inequality and
properties of cosine operators, yields
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđŧ3 (𝑡∗ + ℎ) − đŧ3 (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡∗
ķĩ„Š
= 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫ [đļ (𝑡∗ + ℎ − 𝑠) − đļ (𝑡∗ − 𝑠)] đĩđ‘Ĩ (𝑠) 𝑑𝑠
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Š2
𝑡∗ +ℎ
ķĩ„Š
∗
+∫
đļ (𝑡 + ℎ − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Šķĩ„Š 𝑑𝑠
∗
ķĩ„Šķĩ„Šđ‘ˆ
𝑡
𝑡∗
+ 2∫
𝑡∗ +ℎ
𝑡∗
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđļ (𝑡∗ + ℎ − 𝑠) đĩđ‘Ĩ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠
𝑡∗
ķĩ„Š2
ķĩ„Š
≤ 2𝑡∗ ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Š[đļ (𝑡∗ + ℎ − 𝑠) − đļ (𝑡∗ − 𝑠)] đĩđ‘Ĩ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠
0
2
+ 2ℎ(𝑀∗ ) 𝑀đĩ2 ‖đ‘Ĩ‖2đļ.
(ii) If (A8)–(A10) are satisfied, then (27) where the forcing
term is given by (32) has a unique mild solution on
[0, 𝑇] provided that
1/2 2
(51)
0
Moreover, for every 𝑝 ≥ 1, estimate (49a) holds.
√2𝑀∗ [[2𝑇𝑀𝐹 |𝑎|L2 (1 + 𝑇3 𝑀𝑘 )
2
as |ℎ| ķŗ¨€→ 0,
𝑖=1
ķĩ„Š
ķĩ„Š2
≤ 2𝑡 ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Š[đļ (𝑡∗ + ℎ − 𝑠) − đļ (𝑡∗ − 𝑠)] đĩđ‘Ĩ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠
(i) If (A6) and (A7) are satisfied, then (27) where the
forcing term is given by (33) has a unique mild solution
on [0,T] provided that
2
2
ķĩ„Š
ķĩ„Š2
∑𝐸ķĩ„Šķĩ„Šķĩ„Šđŧ𝑖 (𝑡∗ + ℎ) − đŧ𝑖 (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ķŗ¨€→ 0
∗
Corollary 14. Assume that (A11) and (A13)–(A16) hold.
√2𝑀∗ [2𝑇(𝑇𝑀𝐹 𝑀đĩ + 𝑀𝐹 ) + 𝑀đĩ2 ]
0
1
Proof of Lemma 7. Let 𝜇 ∈ C𝜆2 be fixed and consider the
solution map Φ defined in (39).
One can see from the discussion in Section 2 and the
properties of đ‘Ĩ that for any đ‘Ĩ ∈ đļ([0, 𝑇]; 𝑈), Φ(đ‘Ĩ)(𝑡) is a welldefined stochastic process, for each 0 ≤ 𝑡 ≤ 𝑇. In order to
verify the L2 -continuity of Φ on [0, 𝑇], let 𝑧 ∈ đļ([0, 𝑇]; 𝑈)
and consider 0 ≤ 𝑡∗ ≤ 𝑇 and |ℎ| sufficiently small so that all
terms are well defined. Observe that
ķĩ„Š
ķĩ„Š
ķĩ„Š2
ķĩ„Š2
𝐸ķĩ„Šķĩ„Šķĩ„ŠΦ (𝑧) (𝑡∗ + ℎ) − Φ (𝑧) (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 5 ∑ķĩ„Šķĩ„Šķĩ„Šđŧ𝑖 (𝑡∗ + ℎ) − đŧ𝑖 (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ.
= đŧ1 (𝑡) + đŧ2 (𝑡) + đŧ3 (𝑡) + đŧ5 (𝑡)
𝑡
5. Proofs of Main Results
𝑇1/2 < 1.
(49c)
Moreover, for every 𝑝 ≥ 1, estimate (49a) holds.
Finally, we remark that an approximation scheme in the
spirit of Theorem 10 can be established for (27) by making
the natural modifications to (42) and hypotheses (A20) and
(A21).
(52)
The strong continuity of đļ(𝑡), along with the Lebesgue
dominated convergence theorem, implies that the right side
of (52) goes to 0 as |ℎ| → 0. Regarding the fourth term of
(50), similar computations involving (A11) and (A12) yield the
estimate
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđŧ4 (𝑡∗ + ℎ) − đŧ4 (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑡∗
ķĩ„Š
≤ 2𝑇 ∫ 𝐸 ķĩ„Šķĩ„Šķĩ„Š[𝑆 (𝑡∗ + ℎ − 𝑠) − 𝑆 (𝑡∗ − 𝑠)]
0
ķĩ„Š2
× đ‘“ (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
2 ķĩ„Š
ķĩ„Š2
+ 2ℎ2 (𝑀∗ ) 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘“ (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ.
(53)
International Journal of Stochastic Analysis
9
𝑚 𝑛−1
Since
= 𝐸 ⟨ ∑ ∑ 𝑆ℎ 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 )) ,
ķĩ„Š2
ķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘“ (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝑀𝑓2 [1 + ‖đ‘Ĩ‖2đļ + sup ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđœ†2 ]
0≤𝑠≤𝑇
(54)
𝑗=1 𝑘=0
𝑚 𝑛−1
∑ ∑ 𝑆ℎ 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ))⟩
𝑗=1 𝑘=0
and the right side is independent of ℎ, it follows immediately
from (A11) that the right side of (53) goes to 0 as |ℎ| → 0.
It remains to show that đŧ3 (𝑡∗ +ℎ)−đŧ3 (𝑡∗ ) → 0 as |ℎ| → 0.
Observe that
đŧ5 (𝑡∗ + ℎ) − đŧ5 (𝑡∗ )
𝑡
𝑚 𝑛−1
ķĩ„Š ķĩ„Š2
≤ ∑ ∑ ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘†â„Ž ķĩ„Šķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾2 𝐸
𝑗=1 𝑘=0
× (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ) , đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ))
R
ķĩ„Šķĩ„Š 𝑡∗ +ℎ
ķĩ„Š
= 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫
𝑆 (𝑡∗ + ℎ − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠)
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Š2
𝑡∗
ķĩ„Š
− ∫ 𝑆 (𝑡∗ − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘ˆ
0
ķĩ„Šķĩ„Š ∞ 𝑡∗ +ℎ
ķĩ„Šķĩ„Š
= 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ∑ ∫
𝑆 (𝑡∗ + ℎ − 𝑠) 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗đģ (𝑠)
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Š
ķĩ„Šķĩ„Š2
∞ 𝑡∗
ķĩ„Šķĩ„Š
∗
đģ
− ∑ ∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
𝑗=1 0
ķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š ∞ 𝑡∗ +ℎ
ķĩ„Šķĩ„Š
= 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ∑ ∫
𝑆 (𝑡∗ + ℎ − 𝑠) 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗đģ (𝑠)
ķĩ„Šķĩ„Šđ‘—=1 𝑡∗
ķĩ„Š
∞
𝑈
2
ķĩ„Š ķĩ„Š2
≤ sup ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘†â„Ž ķĩ„Šķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾2 𝐸(đ›Ŋ𝑗đģ)
0≤𝑠≤𝑡∗
𝑚
ķĩ„Š ķĩ„Š2
2đģ
= sup ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘†â„Ž ķĩ„Šķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾2 (𝑡∗ ) ∑ ]𝑗 ,
0≤𝑠≤𝑡∗
𝑗=1
(56)
where 𝑆ℎ = 𝑆(𝑡∗ − 𝑡𝑘 + ℎ) − 𝑆(𝑡∗ − 𝑡𝑘 ). Hence,
(55)
∗
+ ∑ ∫ [𝑆 (𝑡∗ + ℎ − 𝑠) − 𝑆 (𝑡∗ − 𝑠)]
𝑗=1 0
ķĩ„Šķĩ„Š2
ķĩ„Š
đģ
× đ‘” (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠) ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š ∞ 𝑡∗ +ℎ
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
∗
đģ
ķĩ„Š
≤ 2𝐸 [ķĩ„Šķĩ„Šķĩ„Š ∑ ∫
𝑆 (𝑡 + ℎ − 𝑠) 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Š
𝑡∗
ķĩ„Šđ‘ˆ
[ķĩ„Šķĩ„Šđ‘—=1
ķĩ„Šķĩ„Š ∞ 𝑡∗
ķĩ„Šķĩ„Š
+ ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ∑ ∫ [𝑆 (𝑡∗ + ℎ − 𝑠) − 𝑆 (𝑡∗ − 𝑠)]
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Š
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š
đģ
× đ‘” (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠) ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ] .
ķĩ„Šķĩ„Š
ķĩ„Šđ‘ˆ ]
For the moment, assume that 𝑔 is a simple function as defined
in (16). Observe that for 𝑚 ∈ N, arguing as in Lemma 6 of [17]
yields
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑚 𝑡∗
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
∗
∗
đģ
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Š∑ ∫ [𝑆 (𝑡 + ℎ − 𝑠) − 𝑆 (𝑡 − 𝑠)] 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šđ‘ˆ
ķĩ„Š
2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šķĩ„Š 𝑚 𝑛−1
ķĩ„Šķĩ„Š
ķĩ„Š
= 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š ∑ ∑ 𝑆ℎ 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ))ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 𝑘=0
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š ∞ 𝑡∗
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
∗
∗
đģ
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Š∑ ∫ [𝑆 (𝑡 + ℎ − 𝑠) − 𝑆 (𝑡 − 𝑠)] 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šđ‘ˆ
ķĩ„Š
ķĩ„Šķĩ„Š 𝑚 𝑡∗
ķĩ„Šķĩ„Š
≤ lim 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∑ ∫ [𝑆 (𝑡∗ + ℎ − 𝑠) − 𝑆 (𝑡∗ − 𝑠)]
𝑚→∞ ķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š
đģ
(57)
×𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠) ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šđ‘ˆ
𝑚
ķĩ„Š ķĩ„Š2
2đģ
≤ lim sup ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘†â„Ž ķĩ„Šķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾2 (𝑡∗ ) ∑ ]𝑗
𝑚→∞
0≤𝑠≤𝑡∗
𝑗=1
∞
ķĩ„Š ķĩ„Š2
2đģ
= sup ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘†â„Ž ķĩ„Šķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾2 (𝑡∗ ) ∑ ]𝑗 ,
0≤𝑠≤𝑡∗
𝑗=1
and the right side of (57) goes to 0 as |ℎ| → 0. Next, observe
that
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š ∞ 𝑡∗ +ℎ
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
∗
đģ
ķĩ„Š
𝑆 (𝑡 + ℎ − 𝑠) 𝑔 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š ∑ ∫
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 𝑡∗
ķĩ„Šđ‘ˆ
ķĩ„Š
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Šķĩ„Š ∞ ℎ
ķĩ„Šķĩ„Š
ķĩ„Š
= 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∑ ∫ 𝑆 (đ‘ĸ) 𝑔 (𝑡∗ + ℎ − đ‘ĸ) 𝑒𝑗 𝑑đ›Ŋ𝑗đģ (𝑡∗ + ℎ − đ‘ĸ)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š .
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šđ‘ˆ
(58)
2
Using the property 𝐸(đ›Ŋ𝑗đģ(𝑠) − đ›Ŋ𝑗đģ(𝑡)) = |𝑡 − 𝑠|2đģ]𝑗 with 𝑠 =
𝑡∗ + ℎ and 𝑡 = 𝑡∗ , we can argue as above to conclude that
the right side of (58) goes to 0 as |ℎ| → 0. Consequently,
đŧ3 (𝑡∗ + ℎ) − đŧ3 (𝑡∗ ) → 0 as |ℎ| → 0 when 𝑔 is a simple
function. Since the set of all such simple functions is dense in
L(𝑉, 𝑈), a standard density argument can be used to extend
this conclusion to a general bounded, measurable function 𝑔.
This establishes the L2 -continuity of Φ.
10
International Journal of Stochastic Analysis
Finally, we assert that Φ(đļ([0, 𝑇]; 𝑈)) ⊂ đļ([0, 𝑇]; 𝑈).
Indeed, the necessary estimates can be established as above,
and when used in conjunction with Lemma 6, one can
readily verify that sup0≤𝑡≤𝑇 𝐸‖Φ(đ‘Ĩ)(𝑡)‖2𝑈 < ∞, for any đ‘Ĩ ∈
đļ([0, 𝑇]; 𝑈). Thus, we conclude that Φ is well defined, and
the proof of Lemma 7 is complete.
Proof of Theorem 8. Let 𝜇 ∈ C𝜆2 be fixed and consider the
operator as Φ defined in (39). We know that Φ is well defined
and L2 -continuous from Lemma 7. We now prove that Φ has
a unique fixed point in đļ([0, 𝑇]; 𝑈). Indeed, for any đ‘Ĩ, đ‘Ļ ∈
đļ([0, 𝑇]; 𝑈), (39) implies that
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š(Φđ‘Ĩ) (𝑡) − (Φđ‘Ļ) (𝑡)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑡
2
ķĩ„Š2
ķĩ„Š
≤ 2𝑇(𝑀∗ ) (𝑀đĩ2 + 𝑀𝑓2 ) ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ (𝑠) − đ‘Ļ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
(59)
and hence
𝜌 (L (đ‘Ĩ𝜇 (𝑡 + ℎ)) , L (đ‘Ĩ𝜇 (𝑡)))
= sup ∫ 𝜑 (đ‘Ĩ) (L (đ‘Ĩ𝜇 (𝑡+ℎ))−L (đ‘Ĩ𝜇 (𝑡))) (𝑑đ‘Ĩ) ķŗ¨€→ 0
‖𝜑‖𝜆2 ≤1 𝑈
as |ℎ| ķŗ¨€→ 0,
(63)
for any 0 ≤ 𝑡 ≤ 𝑇. Thus, 𝑡 ķŗ¨ƒ→ L(đ‘Ĩ𝜇 (𝑡)) is a continuous map,
so that L(đ‘Ĩ𝜇 ) ∈ C𝜆2 , thereby showing that Ψ is well defined.
In order to show that Ψ has a unique fixed point in C𝜆2 , let
𝜇, ] ∈ C𝜆2 and let đ‘Ĩ𝜇 , đ‘Ĩ] be the corresponding mild solutions
of (26). A standard computation yields
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡) − đ‘Ĩ] (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑡
ķĩ„Š2
ķĩ„Š
2
≤ 2𝑇(𝑀∗ ) (𝑀đĩ2 + 𝑀𝑓2 ) ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠) − đ‘Ĩ] (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
𝑡
ķĩ„Š2
ķĩ„Š
= đ›ŋ ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ (𝑠) − đ‘Ļ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ ,
0
∗ 2
(𝑀đĩ2
2
where đ›ŋ = 2𝑇(𝑀 )
For any natural number 𝑛, it
follows from successive iteration of (59) that
𝑇đ›ŋ𝑛 ķĩ„Šķĩ„Š
ķĩ„Š2
ķĩ„Šķĩ„Š 𝑛
𝑛 ķĩ„Š2
ķĩ„Šķĩ„ŠΦ đ‘Ĩ − Φ đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ ≤
ķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ.
𝑛!
(60)
Since (𝑇đ›ŋ𝑛 /𝑛!) < 1 for large enough values of 𝑛, we can
conclude from (60) that Φ𝑛 is a strict contraction on đļ and so
for a given 𝑇 > 0, Φ has a unique fixed point đ‘Ĩ𝜇 ∈ đļ (by the
Banach contraction mapping principle), and this coincides
with a mild solution of (26) as desired.
To complete the proof, we must show that 𝜇 is, in fact, the
probability law of đ‘Ĩ𝜇 . To this end, let L(đ‘Ĩ𝜇 ) = {L(đ‘Ĩ𝜇 (𝑡)) : 𝑡 ∈
[0, 𝑇]} represent the probability law of đ‘Ĩ𝜇 and define the map
Ψ : C𝜆2 → C𝜆2 by Ψ(𝜇) = L(đ‘Ĩ𝜇 ). It is not difficult to see that
L(đ‘Ĩ𝜇 (𝑡)) ∈ ℘𝜆2 (𝑈), for all 𝑡 ∈ [0, 𝑇] since đ‘Ĩ𝜇 ∈ đļ([0, 𝑇]; 𝑈).
In order to verify the continuity of the map 𝑡 ķŗ¨ƒ→ L(đ‘Ĩ𝜇 (𝑡)),
we first comment that an argument similar to the one used
to establish Lemma 7 can be used to show that for sufficiently
small |ℎ| > 0
ķĩ„Š2
ķĩ„Š
lim 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡 + ℎ) − đ‘Ĩ𝜇 (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ = 0,
ℎ→0 ķĩ„Š
0
0 ≤ 𝑡 ≤ 𝑇.
(64)
Observe that 𝜌2 (𝜇(𝑠), ](𝑠)) ≤ 𝐷𝑇2 (𝜇, ]). Thus, continuing the
inequality in (64) yields
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡) − đ‘Ĩ] (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑡
ķĩ„Š2
ķĩ„Š
2
≤ 2𝑇(𝑀∗ ) (𝑀đĩ2 + 𝑀𝑓2 ) ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠) − đ‘Ĩ] (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘  (65)
0
2
+ 5𝑇2 (𝑀∗ ) 𝑀𝑓2 𝐷𝑇2 (𝜇, ]) ,
0 ≤ 𝑡 ≤ 𝑇.
Applying Gronwall’s lemma now yields
ķĩ„Š
ķĩ„Š2
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡) − đ‘Ĩ] (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
2
2
≤ 5𝑇2 (𝑀∗ ) 𝑀𝑓2 exp (2𝑇(𝑀∗ ) (𝑀đĩ2 + 𝑀𝑓2 )) 𝐷𝑇2 (𝜇, ])
= 𝜍 (𝑇) 𝐷𝑇2 (𝜇, ]) ,
0 ≤ 𝑡 ≤ 𝑇.
(66)
(61)
∀ 0 ≤ 𝑡 ≤ 𝑇.
Consequently, since for all 𝑡 ∈ [0, 𝑇] and 𝜑 ∈ C𝜆2 , it is the
case that
ķĩ„¨ķĩ„¨
ķĩ„¨ķĩ„¨
ķĩ„¨
ķĩ„¨ķĩ„¨
ķĩ„¨ķĩ„¨∫ 𝜑 (đ‘Ĩ) (L (đ‘Ĩ𝜇 (𝑡 + ℎ)) − L (đ‘Ĩ𝜇 (𝑡))) (𝑑đ‘Ĩ)ķĩ„¨ķĩ„¨ķĩ„¨
ķĩ„¨ķĩ„¨
ķĩ„¨ķĩ„¨ 𝑈
ķĩ„¨
ķĩ„¨
= ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨đ¸ [𝜑 (đ‘Ĩ𝜇 (𝑡 + ℎ)) − 𝜑 (đ‘Ĩ𝜇 (𝑡))]ķĩ„¨ķĩ„¨ķĩ„¨ķĩ„¨
ķĩ„Š
ķĩ„Š
ķĩ„Š ķĩ„Š
≤ ķĩ„Šķĩ„Šķĩ„Šđœ‘ķĩ„Šķĩ„Šķĩ„Šđœ†2 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡 + ℎ) − đ‘Ĩ𝜇 (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ,
𝑡
+ 2𝑇(𝑀∗ ) 𝑀𝑓2 ∫ 𝜌2 (𝜇 (𝑠) , ] (𝑠)) 𝑑𝑠,
+ 𝑀𝑓2 ).
We can choose 0 ≤ 𝑇 ≤ 𝑇 to ensure 𝜍(𝑇) < 1, so that taking
the supremum above yields
ķĩ„Šķĩ„Š
ķĩ„Š2
2
ķĩ„Šķĩ„Šđ‘Ĩ𝜇 − đ‘Ĩ] ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šđļ([0,𝑇];𝑈) ≤ 𝜍 (𝑇) 𝐷𝑇 (𝜇, ]) .
(67)
Since
(62)
ķĩ„Š
ķĩ„Š
𝜌 (L (đ‘Ĩ𝜇 (𝑡)) , L (đ‘Ĩ] (𝑡))) ≤ 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡) − đ‘Ĩ] (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ,
∀ 0 ≤ 𝑡 ≤ 𝑇,
(68)
International Journal of Stochastic Analysis
11
𝑡
ķĩ„Š2𝑝 ķĩ„Š
ķĩ„Š
ķĩ„Š2𝑝
+ đļ𝑇∗ 2𝑝 ∫ (𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđœ†2 ) 𝑑𝑠
0
(which follows directly from (62) and (63)), we have
ķĩ„Š2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„ŠΨ (𝜇) − Ψ (])ķĩ„Šķĩ„Šķĩ„Šđœ†2
+
= 𝐷𝑇2 (Ψ (𝜇) , Ψ (]))
ķĩ„Š2
ķĩ„Š
≤ sup 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡) − đ‘Ĩ] (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑡∈[0,𝑇]
2𝑝
2𝑝
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
≤ 10𝑝 (𝑀∗ ) [𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ (1 + 𝑀đĩ )
[
so that Ψ is a strict contraction on C𝜆2 ([0, 𝑇]; (℘𝜆2 (𝑈), 𝜌)).
Thus, (26) has a unique mild solution on [0, 𝑇] with probability distribution 𝜇 ∈ C𝜆2 ([0, 𝑇]; (℘𝜆2 (𝑈), 𝜌)). The solution can
then be extended, by continuity, to the entire interval [0, 𝑇]
in finitely many steps, thereby completing the proof of the
existence-uniqueness portion of the theorem.
Next, we establish the boundedness of 𝑝th moments of
the mild solutions of (26). Let 𝑝 ≥ 1 and observe that the
standard computations yield, for all 0 ≤ 𝑡 ≤ 𝑇,
ķĩ„Š2𝑝
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑡)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
+
∞
𝑝
đļ𝑇 ( ∑]𝑗 )
𝑗=1
𝑝
)]
]
𝑡
ķĩ„Š 2𝑝
ķĩ„Š
2𝑝
𝑝
+ 10𝑝 (𝑀∗ ) (𝑇𝑝 𝑀đĩ + đļ𝑇∗ 2𝑝 ) ∫ 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠.
0
(71)
Therefore, an application of Gronwall’s lemma, followed
by taking the supremum over [0, 𝑇], yields the desired result.
This completes the proof of Theorem 8.
Proof of Theorem 10. We estimate each term of the representation formula for 𝐸‖đ‘Ĩ𝜀 (𝑡) − 𝑧(𝑡)‖2𝑈 separately. We begin
by observing that (A18) and (A19), together, guarantee the
existence of positive constants 𝐾𝑖 and positive functions đ›ŧ𝑖 (𝜀)
(which decrease to zero as 𝜀 → 0+ ) (𝑖 = 1, 2, 3) such that, for
sufficiently small 𝜀,
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2𝑝
ķĩ„Š
ķĩ„Š
+ 𝐸ķĩ„Šķĩ„Šķĩ„Š∫ đļ (𝑡 − 𝑠) đĩđ‘Ĩ (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Šķĩ„Š 2𝑝
ķĩ„Š
+ 𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) 𝑓 (𝑠, đ‘Ĩ (𝑠) , 𝜇 (𝑠)) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2𝑝
ķĩ„Š
ķĩ„Š
+ 𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) 𝑔 (𝑠) 𝑑đ›Ŋđģ (𝑠)ķĩ„Šķĩ„Šķĩ„Š ]
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘†đœ€ (𝑡) đĩ𝑧0 − 𝑆 (𝑡) đĩ𝑧0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝐾1 đ›ŧ1 (𝜀) ,
ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘†đœ€ (𝑡) 𝑧0 − 𝑆 (𝑡) 𝑧0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝐾2 đ›ŧ2 (𝜀) ,
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Šđļ𝜀 (𝑡) 𝑧0 − đļ (𝑡) 𝑧0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝐾3 đ›ŧ3 (𝜀) ,
(70)
2𝑝
2𝑝
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
≤ 10𝑝 (𝑀∗ ) [𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ (1 + 𝑀đĩ )
[
(72)
for all 0 ≤ 𝑡 ≤ 𝑇.
Next, we estimate
𝑝
ķĩ„Š2
ķĩ„Š
+ (𝑇𝑀đĩ ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ )
0
𝑝
𝑡
+
ķĩ„Š2𝑝
ķĩ„Š
đļ𝑇∗ 2𝑝 𝑇 ( sup ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđœ†2
0≤𝑠≤𝑇
Proof of Proposition 9. Computations similar to those used
leading to the contractivity of the solution map in Theorem 8
can be used, along with Gronwall’s lemma, to establish this
result. The details are omitted.
ķĩ„Š2𝑝
ķĩ„Š2𝑝
ķĩ„Š
ķĩ„Š
≤ 5𝑝 [𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘† (𝑡) đ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Š(đļ (𝑡) − 𝑆 (𝑡) đĩ) đ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Š2 ķĩ„Š
ķĩ„Š
ķĩ„Š2
+ (𝑇 ∫ (𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑠)ķĩ„Šķĩ„Šķĩ„Šđœ†2 ) 𝑑𝑠)
0
𝑝
∞
]
(69)
ķĩ„Š2
ķĩ„Š
= ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 − đ‘Ĩ] ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđļ 0,𝑇 ;𝑈 < 𝜍 (𝑇) 𝐷𝑇2 (𝜇, ]) ,
([ ] )
𝑡
𝑝
∞
𝑝
đļ𝑇 ( ∑ ]𝑗 ) ]
𝑗=1
+ (đļ𝑡 ∑]𝑗 ) ] .
𝑗=1
]
Applying the Jensen and HoĖˆlder inequalities enables us to
continue the string of inequalities above as follows:
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ (đļ𝜀 (𝑡 − 𝑠) đĩ𝜀 đ‘Ĩ𝜀 (𝑠) − đļ (𝑡 − 𝑠) đĩ𝑧 (𝑠)) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š .
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
(73)
To this end, note that
𝑡
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) đĩ𝜀 đ‘Ĩ𝜀 (𝑠) − đļ𝜀 (𝑡 − 𝑠) đĩ𝜀 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
∗ 2
≤ (𝑀 )
𝑡
𝑀đĩ2 ∫
0
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ ,
(74)
and that
2𝑝
2𝑝
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
≤ 10𝑝 (𝑀∗ ) [𝐸 ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ1 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ + 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ0 ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ (1 + 𝑀đĩ )
[
𝑡
ķĩ„Š2𝑝
ķĩ„Š
𝑝
+ 𝑇𝑝 𝑀đĩ ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜇 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠
0
𝑡
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) đĩ𝜀 𝑧 (𝑠) − đļ𝜀 (𝑡 − 𝑠) đĩ𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
𝑡
ķĩ„Š2
ķĩ„Š
≤ (𝑀 ) ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđĩ𝜀 𝑧 (𝑠) − đĩ𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ .
∗ 2
0
(75)
12
International Journal of Stochastic Analysis
Since (A19) guarantees that the right side of (75) goes to zero
as 𝜀 → 0+ (uniformly in 𝑡), we know that there exist 𝐾4 > 0
and positive function đ›ŧ4 (𝜀) as in (72) such that, for sufficiently
small 𝜀 > 0,
𝑡
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) đĩ𝜀 𝑧 (𝑠) − đļ𝜀 (𝑡 −
0
ķĩ„Š2
𝑠) đĩ𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
Note that (A20) guarantees the existence of 𝐾7 > 0 and đ›ŧ7 (𝜀)
(as above) such that, for sufficiently small 𝜀 > 0,
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđš2𝜀 (𝑠, 𝑧 (𝑠)) − 𝐹 (𝑠, 𝑧 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ 𝐾7 đ›ŧ7 (𝜀) ,
≤ 𝐾4 đ›ŧ4 (𝜀) ,
(76)
for all 0 ≤ 𝑡 ≤ 𝑇. Finally, (A18) and (A19) guarantee that the
right side of
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) đĩ𝑧 (𝑠) − đļ (𝑡 − 𝑠) đĩ𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
𝑡
ķĩ„Š2
ķĩ„Š
≤ ∫ ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) − đļ (𝑡 − 𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ¸â€–đĩ𝑧 (𝑠)‖2𝑈𝑑𝑠
0
for all 0 ≤ 𝑡 ≤ 𝑇.
So, we can continue the inequality (82) to conclude that
𝑡
ķĩ„Š
ķĩ„Š2
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑠) [𝐹2𝜀 (𝑠, đ‘Ĩ𝜀 (𝑠)) − 𝐹 (𝑠, 𝑧 (𝑠))]ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
𝑡
(77)
≤
tends to zero as 𝜀 → 0+ , uniformly in 𝑡, so that again there
exist 𝐾5 > 0 and positive function đ›ŧ5 (𝜀) as in (72) such that,
for sufficiently small 𝜀 > 0,
(83)
𝑡
4𝑀𝐹2 ∫
0
(84)
ķĩ„Š2
ķĩ„Š
𝐸 ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘  + 4𝑇𝐾7 đ›ŧ7 (𝜀) .
Using (82) and (84), together with the HoĖˆlder,
Minkowski, and Young inequalities, yields
𝑡
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđļ𝜀 (𝑡 − 𝑠) đĩ𝑧 (𝑠) − đļ (𝑡 − 𝑠) đĩ𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘  ≤ 𝐾5 đ›ŧ5 (𝜀) ,
0
(78)
for all 0 ≤ 𝑡 ≤ 𝑇. Thus, using (75)–(78), together with standard inequalities, in (74) yields
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ (đļ𝜀 (𝑡 − 𝑠) đĩ𝜀 đ‘Ĩ𝜀 (𝑠) − đļ (𝑡 − 𝑠) đĩ𝑧 (𝑠)) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
𝑡
2
ķĩ„Š2
ķĩ„Š
≤ 16𝑇1/2 [(𝑀∗ ) 𝑀đĩ2 ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ ]
0
(79)
+ 16𝑇1/2 (𝐾4 đ›ŧ4 (𝜀) + 𝐾5 đ›ŧ5 (𝜀)) .
Next, we estimate
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ (𝑆𝜀 (𝑡 − 𝑠) 𝐹2𝜀 (𝑠, đ‘Ĩ𝜀 (𝑠)) − 𝑆 (𝑡 − 𝑠) 𝐹 (𝑠, 𝑧 (𝑠))) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š .
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
(80)
The continuity of 𝐹, together with (A20), enables us to infer
the existence of 𝐾6 > 0 and đ›ŧ6 (𝜀) (as above) such that, for
sufficiently small 𝜀 > 0,
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑡
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ (𝑆𝜀 (𝑡 − 𝑠) 𝐹2𝜀 (𝑠, đ‘Ĩ𝜀 (𝑠)) − 𝑆 (𝑡 − 𝑠) 𝐹 (𝑠, 𝑧 (𝑠))) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 0
≤ 4𝑇1/2 [𝐾2 đ›ŧ2 (𝜀) + 4𝑇𝐾3 đ›ŧ3 (𝜀)
𝑡
2
ķĩ„Š2
ķĩ„Š
+ 4(𝑀∗ ) 𝑀𝐹2 ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ ] .
0
(85)
Next, (A21) guarantees the existence of 𝐾8 > 0 and đ›ŧ8 (𝜀) (as
above) such that, for sufficiently small 𝜀 > 0,
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝐹1𝜀 (𝑠, 𝑧) 𝜇𝜀 (𝑠) (𝑑𝑧)ķĩ„Šķĩ„Šķĩ„Š ≤ 𝐾8 đ›ŧ8 (𝜀) ,
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑈
∀ 0 ≤ 𝑠 ≤ 𝑇.
(86)
As such, we have
𝑡
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Š[𝑆𝜀 (𝑡 − 𝑠) − 𝑆 (𝑡 − 𝑠)] 𝐹 (𝑠, 𝑧 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘  ≤ 𝐾6 đ›ŧ6 (𝜀) ,
0
(81)
for all 0 ≤ 𝑡 ≤ 𝑇. Also, observe that Young’s inequality and
(A20), together, imply
𝑡
ķĩ„Š2
ķĩ„Š
∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑠) [𝐹2𝜀 (𝑠, đ‘Ĩ𝜀 (𝑠)) − 𝐹 (𝑠, 𝑧 (𝑠))]ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ 
0
𝑡
(87)
2
𝑡
2
ķĩ„Š2
ķĩ„Š
≤ 4(𝑀∗ ) ∫ [𝑀𝐹2 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š2
𝑡 ķĩ„Š
ķĩ„Šķĩ„Š
2
ķĩ„Š
≤ (𝑀∗ ) 𝑇 ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝐹1𝜀 (𝑠, 𝑧) 𝜇𝜀 (𝑠) (𝑑𝑧)ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Šđ‘ˆ
0 ķĩ„Š 𝑈
≤ (𝑀∗ ) 𝑇2 𝐾8 đ›ŧ8 (𝜀) ,
2
ķĩ„Š
≤ (𝑀∗ ) ∫ 𝐸 ķĩ„Šķĩ„Šķĩ„Šđš2𝜀 (𝑠, đ‘Ĩ𝜀 (𝑠)) − 𝐹2𝜀 (𝑠, 𝑧 (𝑠)) +
0
ķĩ„Š2
+ 𝐹2𝜀 (𝑠, 𝑧 (𝑠)) − 𝐹 (𝑠, 𝑧 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ 𝑑𝑠
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆𝜀 (𝑡 − 𝑠) ∫ 𝐹1𝜀 (𝑠, 𝑧) 𝜇𝜀 (𝑠) (𝑑𝑧) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š 𝑑𝑠
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
𝑈
(82)
for all 0 ≤ 𝑡 ≤ 𝑇.
2
𝑡
It remains to estimate 𝐸‖ ∫0 𝑆𝜀 (𝑡 − 𝑠)𝑔𝜀 (𝑠)𝑑đĩđģ(𝑠)‖𝑈. Observe that (A22) implies the existence of 𝐾9 > 0 and đ›ŧ9 (𝜀) (as
above) such that, for sufficiently small 𝜀 > 0,
0
ķĩ„Š2
ķĩ„Š
+𝐸ķĩ„Šķĩ„Šķĩ„Šđš2𝜀 (𝑠, 𝑧 (𝑠)) − 𝐹 (𝑠, 𝑧 (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ] 𝑑𝑠.
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘”đœ€ (𝑡)ķĩ„Šķĩ„Šķĩ„ŠL(𝑉;𝑈) ≤ 𝐾9 đ›ŧ9 (𝜀) ,
0 ≤ 𝑡 ≤ 𝑇.
(88)
International Journal of Stochastic Analysis
13
First, assume that 𝑔𝜀 is a simple function. Observe that for
𝑚 ∈ N,
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑚 𝑡
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
đģ
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š ∑ ∫ 𝑆𝜀 (𝑡 − 𝑠) 𝑔𝜀 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šđ‘ˆ
ķĩ„Š
Now, using (72)–(90), we conclude that there exist positive constants 𝐾𝑖 (𝑖 = 1, . . . , 9) and positive functions
đ›ŧ𝑖 (𝜀) (𝑖 = 1, . . . , 9) such that
9
2
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑡) − 𝑧 (𝑡)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ ∑𝐾𝑖 đ›ŧ𝑖 (𝜀) + 4(𝑀∗ ) 𝑀𝐹2
𝑖=1
ķĩ„Šķĩ„Š 𝑚 𝑛−1
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
đģ
đģ
ķĩ„Š
= 𝐸ķĩ„Šķĩ„Šķĩ„Š ∑ ∑ 𝑆𝜀 (𝑡 − 𝑡𝑘 ) 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗 (𝑡𝑘+1 ) − đ›Ŋ𝑗 (𝑡𝑘 ))ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 𝑘=0
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šđ‘ˆ
𝑡
ķĩ„Š2
ķĩ„Š
× ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑠) − 𝑧 (𝑠)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đ‘ ,
0
𝑚 𝑛−1
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ𝜀 (𝑡) − 𝑧 (𝑡)ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ ≤ Ψ (𝜀) exp (𝜍𝑡) ,
𝑗=1 𝑘=0
𝑚 𝑛−1
𝑈
𝑚 𝑛−1
𝑗=1 𝑘=0
× (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ) , đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ))
R
2
ķĩ„Š
ķĩ„Š2
≤ sup ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑡𝑘 )ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾9 đ›ŧ9 (𝜀) 𝐸(đ›Ŋ𝑗đģ)
Proof of Lemma 12. Using the discussion in Section 2 and the
properties of đ‘Ĩ, it follows that for any đ‘Ĩ ∈ đļ([0, 𝑇]; 𝑈), Γ(đ‘Ĩ)(𝑡)
is a well-defined stochastic process, for each 0 ≤ 𝑡 ≤ 𝑇.
In order to verify the L2 -continuity of Γ on [0, 𝑇], let 𝑧 ∈
đļ([0, 𝑇]; 𝑈) and consider 0 ≤ 𝑡∗ ≤ 𝑇 and |ℎ| sufficiently
small. Observe that
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„ŠΓ (𝑧) (𝑡∗ + ℎ) − Γ (𝑧) (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
0≤𝑠≤𝑡
5
ķĩ„Š
ķĩ„Š2
≤ 5 ∑ 𝐸ķĩ„Šķĩ„Šķĩ„Šđŧ𝑖 (𝑡∗ + ℎ) − đŧ𝑖 (𝑡∗ )ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
𝑚
ķĩ„Š
ķĩ„Š2
= sup ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑡𝑘 )ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾9 đ›ŧ9 (𝜀) 𝑡2đģ ∑ ]𝑗 .
𝑖=1, 𝑖 =Ė¸ 4
𝑗=1
ķĩ„Šķĩ„Š 𝑡∗ +ℎ
ķĩ„Š
+ 𝐸 ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫
𝑆 (𝑡∗ + ℎ − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Š2
𝑡∗
ķĩ„Š
− ∫ 𝑆 (𝑡∗ − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Šķĩ„Š .
ķĩ„Šķĩ„Šđ‘ˆ
0
(89)
Hence,
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆𝜀 (𝑡 − 𝑠) 𝑔𝜀 (𝑠) 𝑑đ›Ŋđģ (𝑠)ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑡∗ +ℎ
ķĩ„Š
ķĩ„Š
𝑆 (𝑡∗ + ℎ − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡∗
2
≤ 2𝑀𝐹2 (𝑀∗ ) ℎ2
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š 𝑚 𝑡
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
đģ
ķĩ„Š
≤ lim 𝐸ķĩ„Šķĩ„Šķĩ„Š∑ ∫ 𝑆𝜀 (𝑡 − 𝑠) 𝑔𝜀 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗 (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
𝑚→∞ ķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šđ‘ˆ
which goes to 0 as |ℎ| → 0. Also,
𝑚
0≤𝑠≤𝑡
(94)
× [1 + ‖đ‘Ĩ‖2đļ([0,𝑇];𝑈) + ‖F (0)‖2đļ([0,𝑇];𝑈) ] ,
ķĩ„Š2
ķĩ„Š
≤ lim ( sup ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑡𝑘 )ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾9 đ›ŧ9 (𝜀) 𝑡2đģ ∑]𝑗 )
∞
(93)
Using HoĖˆlder’s inequality with (A23) yields
ķĩ„Šķĩ„Š2
ķĩ„Šķĩ„Š ∞ 𝑡
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
= 𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∑ ∫ 𝑆𝜀 (𝑡 − 𝑠) 𝑔𝜀 (𝑠) 𝑒𝑗 𝑑đ›Ŋ𝑗đģ (𝑠)ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Šđ‘—=1 0
ķĩ„Šđ‘ˆ
ķĩ„Š
≤ (𝑀𝑆2 𝑡2đģ ∑ ]𝑗 ) 𝐾9 đ›ŧ9 (𝜀) ,
(92)
2
ķĩ„Š
ķĩ„Š2
≤ ∑ ∑ ķĩ„Šķĩ„Šķĩ„Šđ‘†đœ€ (𝑡 − 𝑡𝑘 )ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) 𝐾9 đ›ŧ9 (𝜀) 𝐸
𝑚→∞
0 ≤ 𝑡 ≤ 𝑇,
where 𝜍 = 4(𝑀∗ ) 𝑀𝐹2 and Ψ(𝜀) = ∑9𝑖=1 𝐾𝑖 đ›ŧ𝑖 (𝜀). This completes the proof.
∑ ∑ 𝑆𝜀 (𝑡 − 𝑡𝑘 ) 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 ))⟩
0≤𝑠≤𝑡
(91)
so that an application of Gronwall’s lemma implies
= 𝐸 ⟨ ∑ ∑ 𝑆𝜀 (𝑡 − 𝑡𝑘 ) 𝑔𝑘 𝑒𝑗 (đ›Ŋ𝑗đģ (𝑡𝑘+1 ) − đ›Ŋ𝑗đģ (𝑡𝑘 )) ,
𝑗=1 𝑘=0
0 ≤ 𝑡 ≤ 𝑇,
ķĩ„Šķĩ„Š 𝑡∗
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Šķĩ„Š∫ [𝑆 (ℎ) − đŧ] 𝑆 (𝑡∗ − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
𝑗=1
0 ≤ 𝑡 ≤ 𝑇.
𝑡∗
𝑗=1
(90)
Since the set of all such simple functions is dense in L(𝑉, 𝑈),
a standard density argument can be used to establish estimate
(90) for a general bounded, measurable function 𝑔𝜀 .
ķĩ„Š
ķĩ„Š2
≤ 𝑇 ∫ ķĩ„Šķĩ„Šķĩ„Š[𝑆 (ℎ) − đŧ] 𝑆 (𝑡∗ − 𝑠)ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈)
0
× đ¸â€–F (đ‘Ĩ) (𝑠) ‖2𝑈𝑑𝑠
ķĩ„Š
ķĩ„Š2
≤ 𝑇 sup ķĩ„Šķĩ„Šķĩ„Š[𝑆 (ℎ) − đŧ] 𝑆 (𝑡∗ − 𝑠)ķĩ„Šķĩ„Šķĩ„ŠBL(𝑈) ‖F (đ‘Ĩ)‖L2 ,
0≤𝑠≤𝑇
(95)
14
International Journal of Stochastic Analysis
and the right side goes to 0 as |ℎ| → 0 here as well. Successive
applications of HoĖˆlder’s inequality yields
ķĩ„Šķĩ„Š 2 1/2
ķĩ„Šķĩ„Š 𝑡
ķĩ„Š
ķĩ„Šķĩ„Š
[𝐸ķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š ]
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 0
𝑇
≤ 𝑇1/2 𝑀∗ [∫ ‖F (đ‘Ĩ) (𝑠)‖2L2 (Ω; 𝑈) 𝑑𝑠 ]
1/2
(96)
0
F (đ‘Ĩ) (𝑡) = ∫ 𝑏 (𝑡, 𝑠) 𝐹0 (𝑠, đ‘Ĩ (𝑠)) 𝑑𝑠 + 𝐹1 (𝑡, đ‘Ĩ (𝑡)) ,
0
Subsequently, an application of (A23), together with Minkowski’s inequality, enables us to continue the string of inequalities in (96) to conclude that
1/2
≤𝑇
(97)
∗
𝑀 [𝑀F ‖đ‘Ĩ‖C + ‖𝐹 (0)‖L2 ] .
Taking the supremum over [0, 𝑇] in (97) then implies that
𝑡
∫0 𝑆(𝑡 − 𝑠)F(đ‘Ĩ)(𝑠)𝑑𝑠 ∈ đļ([0, 𝑇]; 𝑈), for any đ‘Ĩ ∈ đļ([0, 𝑇]; 𝑈).
The other estimates can be established as before, and
when used in conjunction with Lemma 6, they can be used to
verify that sup0≤𝑡≤𝑇 𝐸‖Γ(đ‘Ĩ)(𝑡)‖2𝑈 < ∞, for any đ‘Ĩ ∈ đļ([0, 𝑇];
𝑈). Thus, we conclude that Γ is well defined, and the proof of
Lemma 12 is complete.
Proof of Theorem 13. We know that Γ is well defined and L2 continuous from Lemma 12. We now prove that Γ has a
unique fixed point in đļ([0, 𝑇]; 𝑈). For any đ‘Ĩ, đ‘Ļ ∈ đļ([0, 𝑇; 𝑈])
(Γđ‘Ĩ) (𝑡) − (Γđ‘Ļ) (𝑡)
𝑡
= ∫ đļ (𝑡 − 𝑠) đĩ (đ‘Ĩ (𝑠) − đ‘Ļ (𝑠)) 𝑑𝑠
0
(98)
𝑡
+ ∫ 𝑆 (𝑡 − 𝑠) [F (đ‘Ĩ) (𝑠) − F (đ‘Ļ) (𝑠)] 𝑑𝑠.
0
Squaring both sides, taking the expectation in (98), and
applying Young’s inequality yields
(101)
∀ 0 ≤ 𝑡 ≤ 𝑇.
The Uniform Boundedness Principle with (A6) guarantees
the existence of a positive constant 𝑀𝑏 such that ‖đĩ(𝑡, 𝑠)‖BL ≤
𝑀𝑏 , for all 0 ≤ 𝑡 ≤ 𝑠 ≤ 𝑇. Standard computations involving
properties of expectation and HoĖˆlder’s inequality imply that
for all đ‘Ĩ, đ‘Ļ ∈ đļ([0, 𝑇]; 𝑈),
ķĩ„Š2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„ŠF (đ‘Ĩ) − F (đ‘Ļ)ķĩ„Šķĩ„Šķĩ„ŠL2
𝑇
𝑠
ķĩ„Š
ķĩ„Š2
≤ 2 ∫ [𝑇𝑀𝑏2 ∫ 𝐸ķĩ„Šķĩ„Šķĩ„Šđš0 (𝜏, đ‘Ĩ (𝜏)) − 𝐹0 (𝜏, đ‘Ļ (𝜏))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆđ‘‘đœ
0
0
ķĩ„Š
ķĩ„Š2
+ 𝐸ķĩ„Šķĩ„Šķĩ„Šđš1 (𝑠, đ‘Ĩ (𝑠)) − 𝐹1 (𝑠, đ‘Ļ (𝑠))ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ]
ķĩ„Š
ķĩ„Š
≤ 2𝑇 [𝑇𝑀𝑏 𝑀𝐹0 + 𝑀𝐹1 ] ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ.
(102)
Thus, if we let 𝑀F = 2 𝑇[𝑇𝑀𝑏 𝑀𝐹0 + 𝑀𝐹1 ] in (A23) we can
conclude from Theorem 13 that the IBVP has a unique mild
solution on [0, 𝑇], provided that (49a) holds.
For (ii), using similar reasoning with
𝑀F = 2𝑀𝐹2 𝑇|𝑎|L2 ((0,𝑇)2 ) (1 + 𝑀𝑘 𝑇3 )
1/2
(103)
enables us to invoke Theorem 13 to conclude that the IBVP
has a unique mild solution on [0, 𝑇], provided that (49b)
holds.
6. Application of Theory to
Nonlinear Beam Model
ķĩ„Š2
ķĩ„Š
𝐸ķĩ„Šķĩ„Šķĩ„Š(Γđ‘Ĩ) (𝑡) − (Γđ‘Ļ) (𝑡) ķĩ„Šķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Š2
ķĩ„Š
ķĩ„Š
≤ 2𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) [F (đ‘Ĩ) (𝑠) − F (đ‘Ļ) (𝑠)] 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Šķĩ„Š2
(99)
ķĩ„Š
+ 2𝐸ķĩ„Šķĩ„Šķĩ„Š∫ đļ (𝑡 − 𝑠) đĩ (đ‘Ĩ (𝑠) − đ‘Ļ (𝑠)) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
2
ķĩ„Š2
ķĩ„Š2
ķĩ„Š
ķĩ„Š
≤ 2(𝑀∗ ) 𝑇 [ķĩ„Šķĩ„Šķĩ„ŠF (đ‘Ĩ) − F (đ‘Ļ)ķĩ„Šķĩ„Šķĩ„ŠL2 + 𝑀đĩ2 ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ]
2
ķĩ„Š2
ķĩ„Š
2
+ 𝑀đĩ2 ) ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ,
≤ 2(𝑀∗ ) 𝑇 (𝑀F
We now apply the results proved in Section 5 to the original
nonlinear beam model. Consider the IBVP (4)–(6) equipped
with the nonlinear forcing term (9).
Proposition 15. Assume (A1)–(A5). Then this IBVP has a
unique mild solution 𝑊 ∈ đļ([0, 𝑇]; L2 (0, 𝑇)) with probability distribution 𝜇. Moreover, for every 𝑝 ≥ 1, there exists a
constant đļ𝑝 for which
2𝑝
sup 𝐸‖𝑊 (⋅, 𝑡)‖L2 (0,đŋ)
and so
ķĩ„Š2
ķĩ„Š2
ķĩ„Šķĩ„Š
∗ 2
2
2 ķĩ„Š
ķĩ„Šķĩ„ŠΓđ‘Ĩ − Γđ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ ≤ 2(𝑀 ) 𝑇 (𝑀F + 𝑀đĩ ) ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ.
Proof of Corollary 14. For (i), define F : đļ([0, 𝑇]; 𝑈) →
L2 (0, 𝑇; L2 (Ω; 𝑈)) by
𝑡
≤ 𝑇1/2 𝑀∗ ‖F (đ‘Ĩ)‖L2 .
1/2
ķĩ„Šķĩ„Š 𝑡
ķĩ„Šķĩ„Šķĩ„Š 2
ķĩ„Š
[𝐸ķĩ„Šķĩ„Šķĩ„Š∫ 𝑆 (𝑡 − 𝑠) F (đ‘Ĩ) (𝑠) 𝑑𝑠ķĩ„Šķĩ„Šķĩ„Š ]
ķĩ„Šķĩ„Š 0
ķĩ„Šķĩ„Šđ‘ˆ
Hence, Φ is a strict contraction, provided that
2 + 𝑀2 ) < 1 and so has a unique fixed point
𝑀∗ √2𝑇(𝑀F
đĩ
which coincides with a mild solution of (27). This completes the proof.
0≤𝑡≤𝑇
(100)
ķĩ„Š ķĩ„Š2𝑝
ķĩ„Š ķĩ„Š2𝑝
≤ đļ𝑝 (1 + 𝐸ķĩ„Šķĩ„Šķĩ„Šđœ‰0 ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ) + 𝐸ķĩ„Šķĩ„Šķĩ„Šđœ‰1 ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ) ) .
(104)
International Journal of Stochastic Analysis
15
Proof. We have illustrated in Section 3 that the IBVP (4)–(6)
equipped with (9) can be reformulated as the abstract evolution equation (26). We verify that with those identifications,
(A11)–(A16) are satisfied.
We have already addressed (A11) in Section 3. To see that
(A12) is satisfied, observe that (A1)(i) implies that
Also, invoking (A2)(ii) enables us to see that for all 𝜇, ] ∈
℘𝜆2 (𝑈),
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
𝐹 (𝑡, ⋅, đ‘Ļ) 𝜇 (𝑡, ⋅) (𝑑đ‘Ļ)
ķĩ„Šķĩ„Š∫
ķĩ„Šķĩ„Š L2 (0,đŋ) 4
ķĩ„Šķĩ„Š
ķĩ„Š
𝐹4 (𝑡, ⋅, đ‘Ļ) ] (𝑡, ⋅) (𝑑đ‘Ļ)ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„ŠL2 (0,đŋ)
L2 (0,đŋ)
−∫
ķĩ„Š
ķĩ„Šķĩ„Š
2
ķĩ„Šķĩ„Šđš3 (𝑡, ⋅, đ‘Ĩ (𝑡, ⋅))ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ) ≤ 𝑀𝐹3 [∫ [1 + |đ‘Ĩ (𝑡, 𝑧)|] 𝑑𝑧]
D
≤ 2𝑀𝐹3 [đŋ + ‖đ‘Ĩ (𝑡, ⋅)‖2L2 (0,đŋ) ]
1/2
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
(109)
ķĩ„Š
ķĩ„Š
= ķĩ„Šķĩ„Šķĩ„Š∫
𝐹4 (𝑡, ⋅, đ‘Ļ) (𝜇 (𝑡, ⋅) − ] (𝑡, ⋅)) (𝑑đ‘Ļ)ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š L2 (0,đŋ)
ķĩ„Šķĩ„ŠL2 (0,đŋ)
ķĩ„Š
ķĩ„Š
≤ ķĩ„Šķĩ„Šķĩ„ŠđœŒ (𝜇 (𝑡) , ] (𝑡))ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ)
1/2
≤ 2𝑀𝐹3 [√đŋ + ‖đ‘Ĩ‖đļ([0,𝑇];L2 (0,đŋ)) ]
≤ √đŋ𝜌 (𝜇 (𝑡) , ] (𝑡)) ,
≤ 𝑀𝐹∗3 [1 + ‖đ‘Ĩ‖đļ([0,𝑇];L2 (0,đŋ)) ] ,
(105)
Combining (105) and (108), we see that 𝑓 satisfies (A12)(i)
with
𝑀𝑓1 = 2 ⋅ max {𝑀𝐹4 √đŋ, 𝑀𝐹∗3 } ,
2
for all 0 ≤ 𝑡 ≤ 𝑇 and đ‘Ĩ ∈ đļ([0, 𝑇]; L (0, đŋ)), where
2𝑀𝐹3 √đŋ,
𝑀𝐹∗3 = {
2𝑀𝐹3 ,
(106)
if đŋ ≤ 1.
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Šđš3 (𝑡, ⋅, đ‘Ĩ (𝑡, ⋅)) − 𝐹3 (𝑡, ⋅, đ‘Ļ (𝑡, ⋅))ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ)
1/2
(107)
0
ķĩ„Š
ķĩ„Š
≤ 𝑀𝐹3 ķĩ„Šķĩ„Šķĩ„Šđ‘Ĩ − đ‘Ļķĩ„Šķĩ„Šķĩ„Šđļ([0,𝑇];L2 (0,đŋ)) .
Next, using (A2)(i) together with the HoĖˆlder inequality,
we observe that
ķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š
ķĩ„Š
ķĩ„Šķĩ„Š
𝐹4 (𝑡, ⋅, đ‘Ļ) 𝜇 (𝑡, ⋅) (𝑑đ‘Ļ)ķĩ„Šķĩ„Šķĩ„Š
ķĩ„Šķĩ„Š∫
ķĩ„Šķĩ„ŠL2 (0,đŋ)
ķĩ„Šķĩ„Š L2 (0,đŋ)
2
đŋ
= [∫ [∫
L2 (0,đŋ)
0
0
L2 (0,đŋ)
ķĩ„Šķĩ„Š
ķĩ„Š2
ķĩ„Šķĩ„Šđš4 (𝑡, 𝑧, đ‘Ļ)ķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ) 𝜇 (𝑡, 𝑧) (𝑑đ‘Ļ) 𝑑𝑧]
đŋ
≤ 𝑀𝐹4 [∫ (∫
0
L2 (0,đŋ)
2
ķĩ„Š ķĩ„Š
(1 + ķĩ„Šķĩ„Šķĩ„Šđ‘Ļķĩ„Šķĩ„Šķĩ„ŠL2 (0,đŋ) ) 𝜇 (𝑡, 𝑧) (𝑑đ‘Ļ)) 𝑑𝑧]
1/2
ķĩ„Š
ķĩ„Š
≤ 𝑀𝐹4 √đŋ√ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑡)ķĩ„Šķĩ„Šķĩ„Šđœ†2
ķĩ„Š
ķĩ„Š
≤ 𝑀𝐹4 √đŋ (1 + ķĩ„Šķĩ„Šķĩ„Šđœ‡ (𝑡)ķĩ„Šķĩ„Šķĩ„Šđœ†2 ) ,
(111)
Similar results can be established for the IBVP (4)–(6)
equipped with the nonlinear forcing term described by (32)
or (33) by using Corollary 14 and Theorem 13. We omit the
details.
Finally, we remark that the mild solutions for each of
the IBVPs introduced in Section 1 depend continuously on
the initial data and that the approximation scheme in which
the impact of the stochastic term can be made sufficiently
negligible (cf. Theorem 10) can be applied to each of these
IBVPs.
References
1/2
𝐹4 (𝑡, 𝑧, đ‘Ļ) 𝜇 (𝑡, 𝑧) (𝑑đ‘Ļ)] 𝑑𝑧]
1/2
đŋ
≤ [∫ ∫
𝑀𝑓 = max {𝑀𝐹3 , √đŋ} .
This verifies (A12). Next, (A13), (A15), and (A16) hold by
assumption, and (A14) follows because of the uniform boundedness of {‖𝐴1/2 đ‘Ĩ(𝑡)‖L2 (0,đŋ) : 0 ≤ 𝑡 ≤ 𝑇}. Thus, we can
invoke Theorem 8 to conclude that this IBVP has a unique
mild solution đ‘Ĩ ∈ đļ([0, 𝑇]; L2 (0, đŋ)) with probability law
{𝜇(𝑡, ⋅) : 0 ≤ 𝑡 ≤ 𝑇}.
Also, from (A1)(ii), we obtain
ķĩ„¨2
ķĩ„¨
≤ 𝑀𝐹3 [∫ ķĩ„¨ķĩ„¨ķĩ„¨đ‘Ĩ (𝑡, 𝑧) − đ‘Ļ (𝑡, 𝑧)ķĩ„¨ķĩ„¨ķĩ„¨ 𝑑𝑧]
(110)
and combining (107) and (109), we see that 𝑓 satisfies
(A12)(ii) with
if đŋ > 1,
đŋ
∀ 0 ≤ 𝑡 ≤ 𝑇.
∀ 0 ≤ 𝑡 ≤ 𝑇, 𝜇 ∈ ℘𝜆2 (𝑈) .
(108)
[1] H. E. Lindberg, Little Book of Dynamic Buckling, LCE Science/Software, 2003.
[2] R. W. Dickey, “Free vibrations and dynamic buckling of the
extensible beam,” Journal of Mathematical Analysis and Applications, vol. 29, no. 2, pp. 443–454, 1970.
[3] W. E. Fitzgibbon, “Global existence and boundedess of solutions
to the extensible beam equation,” SIAM Journal on Mathematical Analysis, vol. 13, pp. 739–745, 1982.
[4] J. M. Ball, “Initial-boundary value problems for an extensible
beam,” Journal of Mathematical Analysis and Applications, vol.
42, no. 1, pp. 61–90, 1973.
[5] S. K. Patcheu, “On a global solution and asymptotic behaviour
for the generalized damped extensible beam equation,” Journal
of Differential Equations, vol. 135, no. 2, pp. 299–314, 1997.
16
[6] K. Balachandran and J. Y. Park, “Existence of solutions of a class
of abstract second order nonlinear integrodifferential equations,” Journal of Applied Mathematics and Stochastic Analysis,
vol. 15, no. 2, pp. 115–124, 2002.
[7] D. Kannan and A. T. Bharucha-Reid, “On a stochastic integrodifferential evolution equation of Volterra type,” Journal of
Integral Equations and Applications, vol. 10, pp. 351–379, 1985.
[8] N. I. Mahmudov and M. A. McKibben, “Abstract secondorder damped McKean-Vlasov stochastic evolution equations,”
Stochastic Analysis and Applications, vol. 24, no. 2, pp. 303–328,
2006.
[9] P. J. Chen and M. E. Curtin, “On a theory of heat conduction
involving two temperatures,” Zeitschrift fuĖˆr Angewandte Mathematik und Physik, vol. 19, pp. 614–627, 1968.
[10] D. A. Dawson and J. GaĖˆrtner, “Large deviations for the McKeanVlasov limit for weakly interacting diffusions,” Stochastics, vol.
20, pp. 247–308, 1987.
[11] N. U. Ahmed and X. Ding, “A semilinear Mckean-Vlasov
stochastic evolution equation in Hilbert space,” Stochastic Processes and their Applications, vol. 60, no. 1, pp. 65–85, 1995.
[12] D. N. Keck and M. A. McKibben, “Functional integro-differential stochastic evolution equations in Hilbert space,” Journal
of Applied Mathematics and Stochastic Analysis, vol. 16, no. 2, pp.
141–161, 2003.
[13] M. McKibben, “Second order damped functional stochastic
evolution equations in Hilbert spaces,” Dynamic Systems and
Applications, vol. 12, no. 3-4, pp. 467–488, 2003.
[14] L. Decreusefond and A. S. UĖˆstuĖˆnel, “Stochastic analysis of the
fractional brownian motion,” Potential Analysis, vol. 10, no. 2,
pp. 177–214, 1999.
[15] W. Grecksch and V. V. Anh, “A parabolic stochastic differential
equation with fractional Brownian motion input,” Statistics and
Probability Letters, vol. 41, no. 2, pp. 337–346, 1999.
[16] B. Maslowski and D. Nualart, “Evolution equations driven by
a fractional Brownian motion,” Journal of Functional Analysis,
vol. 202, no. 1, pp. 277–305, 2003.
[17] V. V. Anh and W. Greksch, “A fractional stochastic evolution
equation driven by fractional Brownian motion,” Monte Carlo
Methods and Applications, vol. 9, no. 3, pp. 189–199, 2003.
[18] T. E. Duncan, B. Pasik-Duncan, and B. Maslowski, “Fractional
brownian motion and stochastic equations in hilbert space,”
Stochastics and Dynamics, vol. 2, no. 2, pp. 225–250, 2002.
[19] J. Bochenek, “An abstract nonlinear second-order differential
equation,” Annales Polonici Mathematici, vol. 2, pp. 155–166,
1991.
[20] J. P. Dauer and N. Mahmudov, “Remark on existence result for
second order evolution equations in Banach spaces,” International Journal of Pure and Applied Mathematics, vol. 12, no. 4,
pp. 471–482, 2004.
[21] H. O. Fattorini, Second Order Linear Differential Equations in
Banach Spaces, vol. 108 of North Holland Mathematics Studies,
Elsevier Science, Amsterdam, The Netherlands, 1985.
[22] A. Pazy, Semigroups of Linear Operators and Applications to
Partial Differential Equations, Springer, New York, NY, USA,
1983.
International Journal of Stochastic Analysis
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