Loading [MathJax]/jax/output/SVG/jax.js
Research article

Fully nonlocal stochastic control problems with fractional Brownian motions and Poisson jumps

  • Received: 13 November 2021 Accepted: 03 March 2021 Published: 10 March 2021
  • MSC : 60H15, 35A01, 47H06

  • In this paper, we establish a representation of mild solutions to fully nonlocal stochastic evolution problems. Through the iterative technique and energy estimates, we obtain the existence and uniqueness of mild solution. Furthermore, we prove the existence of optimal control for fully nonlocal stochastic control problems with a non-convex cost function. Two examples are given at the end.

    Citation: Yongqiang Fu, Lixu Yan. Fully nonlocal stochastic control problems with fractional Brownian motions and Poisson jumps[J]. AIMS Mathematics, 2021, 6(5): 5176-5192. doi: 10.3934/math.2021307

    Related Papers:

    [1] Kung-Chi Chen, Kuo-Shing Chen . Pricing green financial options under the mixed fractal Brownian motions with jump diffusion environment. AIMS Mathematics, 2024, 9(8): 21496-21523. doi: 10.3934/math.20241044
    [2] Jiali Wu, Maoning Tang, Qingxin Meng . A stochastic linear-quadratic optimal control problem with jumps in an infinite horizon. AIMS Mathematics, 2023, 8(2): 4042-4078. doi: 10.3934/math.2023202
    [3] Xueqi Wen, Zhi Li . pth moment exponential stability and convergence analysis of semilinear stochastic evolution equations driven by Riemann-Liouville fractional Brownian motion. AIMS Mathematics, 2022, 7(8): 14652-14671. doi: 10.3934/math.2022806
    [4] Noorah Mshary, Hamdy M. Ahmed, Ahmed S. Ghanem . Existence and controllability of nonlinear evolution equation involving Hilfer fractional derivative with noise and impulsive effect via Rosenblatt process and Poisson jumps. AIMS Mathematics, 2024, 9(4): 9746-9769. doi: 10.3934/math.2024477
    [5] Rajesh Dhayal, Muslim Malik, Syed Abbas . Solvability and optimal controls of non-instantaneous impulsive stochastic neutral integro-differential equation driven by fractional Brownian motion. AIMS Mathematics, 2019, 4(3): 663-683. doi: 10.3934/math.2019.3.663
    [6] Kinda Abuasbeh, Ramsha Shafqat, Azmat Ullah Khan Niazi, Muath Awadalla . Nonlocal fuzzy fractional stochastic evolution equations with fractional Brownian motion of order (1,2). AIMS Mathematics, 2022, 7(10): 19344-19358. doi: 10.3934/math.20221062
    [7] Ramkumar Kasinathan, Ravikumar Kasinathan, Dumitru Baleanu, Anguraj Annamalai . Well posedness of second-order impulsive fractional neutral stochastic differential equations. AIMS Mathematics, 2021, 6(9): 9222-9235. doi: 10.3934/math.2021536
    [8] Noorah Mshary, Hamdy M. Ahmed . Discussion on exact null boundary controllability of nonlinear fractional stochastic evolution equations in Hilbert spaces. AIMS Mathematics, 2025, 10(3): 5552-5567. doi: 10.3934/math.2025256
    [9] Weiguo Liu, Yan Jiang, Zhi Li . Rate of convergence of Euler approximation of time-dependent mixed SDEs driven by Brownian motions and fractional Brownian motions. AIMS Mathematics, 2020, 5(3): 2163-2195. doi: 10.3934/math.2020144
    [10] Xiuxian Chen, Zhongyang Sun, Dan Zhu . Mean-variance investment and risk control strategies for a dynamic contagion process with diffusion. AIMS Mathematics, 2024, 9(11): 33062-33086. doi: 10.3934/math.20241580
  • In this paper, we establish a representation of mild solutions to fully nonlocal stochastic evolution problems. Through the iterative technique and energy estimates, we obtain the existence and uniqueness of mild solution. Furthermore, we prove the existence of optimal control for fully nonlocal stochastic control problems with a non-convex cost function. Two examples are given at the end.



    The theory of fractional derivatives is available for describing the tailing phenomena in time and nonlocal appearance in space. These characteristics ensure that fractional derivatives can be widely used in engineering, such as material memory, fluid dynamics and so on, see [1] and reference therein. A fractional Brownian motion (fBm) BH is a centered Guassian process. Especially, it is neither a semi-martingale nor a Markov process when H1/2. Fractional Brownian motions are widely used in modelling of fractal phenomena and stock markets, see [2,3]. Hence, it is significant to study the complex systems in Rd as follows

    {Dc0Dβtu(t,x)+(Δ)α2u(t,x)=f(t,x,u(t,x))+Av(x)+g(t,x)dBH(t)dt+Yh(t,x,u(t,x);y)˜η(dy),t(0,),xRd,u(0,x)=u0(x),xRd, (1.1)

    where parameters β(1/2,1), α(1,2) and H(1/2,1). Dc0Dβt denotes the β-order Caputo fractional derivative and (Δ)α2 is the fractional Laplacian. f and h are nonlinear functions. BH is a fBm with Hurst index H and ˜η is a Poisson martingale measure. Let V be a Hilbert space, vV and AL(V;L2(Rd)). Av is the control term.

    In problem (1.1), there are fractional derivatives in time and space, involving both fBm term and Poisson jump. There are two difficulties to study this type of problem. One is mathematical methods to deal with fBms and Poisson jumps driven time-space fractional stochastic systems. Another is Lp-estimates of the Green functions caused by the inconvenience of fractional calculus. There have been a series of researches on this topic:

    1.The fully nonlocal deterministic problems.

    {Dc0Dβtu(t,x)+(Δ)α2u(t,x)=f(t,x,u(t,x)),t(0,),xRd,u(0,x)=u0,xRd. (1.2)

    Kemppainen, Siljander and Zacher [4] dealt with the representation of solutions and the L2-decay of mild solutions for (1.2) when right hand term is f(t,x). Li, Liu and Wang [5] worked on the situation of f(t,x,u)=(u(Bu)) and obtained rich results.

    2.The stochastic evolution problems with fBms and Poisson jumps, i.e., β=1 and α=2.

    It is well known that the smaller H is, the harder the problem is, since the lack of regularity. Tindel, Tudor and Viens [6] discussed a class of linear stochastic evolution equations driven by fBms and obtained the well-posedness of solutions under Hurst index H(1/2,1) and H(0,1/2), respectively. Caraballo, Garrido-Atienza and Taniguchi [7] investigated the abstract stochastic evolution equations with fBms in Hilbert space and considered the existence of weak solutions when H(1/2,1). Duc et al. [8] discussed the case of H(1/2,1) as well and obtained the exponential stability. Tang and Meng [9] studied stochastic evolution equations with Poisson jumps and obtained the existence and uniqueness of solutions by the Gälerkin approximation under Lipschitz type conditions, in addition, they also considered the optimal control problem of the equations.

    3.The fractional stochastic problems.

    Gu et al. [10] dealt with the problems of space-fractional stochastic reaction diffusion equations. They proved the existence, uniqueness and compactness of solution. Chen, Hu and Nualart [11] studied the time-space fractional stochastic equations with Wiener processes and obtained the well-posedness of solutions. Recently, Ahmed et al. considered several stochastic integrodifferential equations with fractional power operators. For examples, in [12] and [13] stochastic integrodifferential equations with Hilfer fractional derivatives driven by fBms were discussed. Ahmed [14] considered an abstract nonlocal stochastic integrodifferential system with Caputo fractional derivatives driven by fBms and Poisson jumps and obtained the existence of mild solutions by fixed point theory provided the order of Caputo fractional derivative ranges from 1 to 2.

    We use iterative technique to deal with problem (1.1), where the order of Caputo fractional derivative β(1/2,1), under Lipschitz nonlinearities as well as proper assumptions on noise terms.

    We are also interested in optimal control problems. Optimal control is a fundamental topic in control theory. Due to wildly applications in engineering and medical science, optimal control problems are studied by a lot of researchers, see [9,15,16] and references therein. The convexity of cost functions and the convexity of the set of control actions are two important factors in optimal control problems. Recently, Benner and Trautwein [17] discussed the existence of optimal control of stochastic heat equations with a convex cost function. Fuhrman, Hu and Tessitore [18] proved the stochastic maximum principle of stochastic partial differential equations and applied it to obtain the optimal control. Durga and Muthukumar [19] investigated optimal control problem of fractional stochastic equations with Poisson jumps. Inspired by the papers above, we study the optimal control problem of (1.1) with an abstract non-convex cost function. Under suitable assumptions on cost function and the compactness of the admissible set, we prove the existence of optimal control.

    The paper is organized as follows. In Section 2, we present basic notions and relative results. In Section 3, we show the representation of mild solutions and prove the existence and uniqueness by iterative technique. In Section 4, we discuss the optimal control problem. In Section 5, we give two applications. In Section 6, we give conclusions.

    The β-order Caputo fractional derivative of u is defined by

    Dc0Dβtu(t)=1Γ(1β)t0(ts)βddsu(s)ds,

    where Γ is Gamma function. The fractional Laplacian is defined by

    (Δ)α2u(x)=F1(|ξ|αFu(ξ))(x),α(0,2),

    where F and F1 denote the Fourier transform and the inverse Fourier transform, respectively. Let the function space

    Wα2,2(Rd)={uL2(Rd),F1[(1+|ξ|2)α2Fu]L2(Rd)}

    be endowed with the norm

    uWα2,2=F1[(1+|ξ|2)α2Fu]2,

    where 2 denotes the norm of L2(Rd).

    Remark 1. Fractional Laplacian can also be defined through singular integral

    (Δ)α2φ(x)=C(d,α)P.V.Rdφ(x)φ(y)|xy|d+αdy,xRd,

    where α(0,2). These two approaches are equivalent on some levels, see [20].

    The bi-parameters Mittag-Leffler function is defined by

    Mk,l(z)=n=1znΓ(nk+l),zC.

    Function Mk,l(z) is the extension of exponential function and M1,1(z)=ez. Let

    FΦ(t,z)=Mβ,1(|ξ|αtβ),FJ(t,z)=Mβ,β(|ξ|αtβ)

    and

    Rβα(t)F=Φ(t,z)F,Sβα(t)F=tβ1J(t,z)FΨ(t,z)F,

    where α(1,2],β(0,1) and denotes convolution. In fact Rβα(t)u0 is a formal solution of problem

    {Dc0Dβtu(t,x)+(Δ)α2u(t,x)=0,t(0,),xRd,u(0,x)=u0(x),xRd.

    Therefore, by Laplace transform and Duhamel's formula, the mild solution of (1.2) can be expressed as

    u=Rβα(t)u0+t0Sβα(ts)f(s,u)ds=RdΦ(t,xz)u0(z)dz+t0RdΨ(ts,xz)f(s,z,u(s,z))dzds,

    if each integral is well defined.

    We recall the continuity and Lp estimates of operators Φ,Ψ,Rβα and Sβα. Let

    χ1={ddα,d>α,,otherwise,χ2={dd2α,d>2α,,otherwise,

    for β(0,1) and α(1,2).

    Lemma 1. [5] Assume β(0,1) and α(1,2).

    1. Let p[1,χ1). Then Φ(t,x)C((0,),Lp(Rd)) and there exists a real number C>0 such that

    Φ(t)pCtdβα(11p).

    2. Let p[1,χ2). Then Ψ(t,x)C((0,),Lp(Rd)) and there exists a real number C>0 such that

    Ψ(t)pCtdβα(11p)+β1.

    Let q[1,) and

    κ1={qddqα,d>qα,,otherwise,κ2={qdd2qα,d>2qα,,otherwise.

    Lemma 2. [5] Assume β(0,1),α(1,2).

    1. Let p[1,κ1). Then

    Rβα(t)upCtβdα(1q1p)uq.

    2. Let p[1,κ2). Then

    Sβα(t)upCtβdα(1q1p)+β1uq.

    3. Let r[1,) and uLr(Rd). Then the mapping

    tRβα(t)uC([0,),Lr(Rd)).

    Let (Ω,F,P) denote a complete probability space endowed with a usual filtration {Ft}0t< and LFt2([0,T];L2(Rd)) be the space of all Ft-adapted random processes such that

    u2LFt2=E[supt[0,T]u(t)22]<.

    Denote L2(Ω;L2(Rd)) the family of all random variables such that

    ζ2L2(Ω;L2(Rd))=Eζ22<.

    We introduce a general concept of fBms. Let BH(t) be a fBm on (Ω,F,P) and U be a separable Hilbert space with an orthonormal basis {ei}i=1. The space of continuous bounded linear operators on U is denoted by L(U). Let QL(U) be a symmetric operator and {λi}i=1 be a sequence of eigenvalues of Q. Further assume that Q is nonnegative and trQ<. Then the Q-fBm on U is defined by

    BHQ(t)=i=1λieiβHi(t),

    where {βHi(t)}t[0,T] is a sequence of 1-dimensional fBms. BHQ is a Gaussian process and its mean and covariance are

    EBHQ(t),yU=0andEBHQ(t),xUBHQ(s),yU=ρ(t,s)Qx,yU,

    where t,s[0,T], x,yU and ρ(t,s) is the covariance operator,

    ρ(t,s)=E[βH(t)βH(s)]=12(t2H+s2H|ts|2H).

    Besides, BHQ becomes standard Brownian motion when Q is an identical operator and Hurst index H=1/2.

    We recall relative results of stochastic calculus with respect to fBms, see [6,21] and reference therein. First, we call ψ an element of Hilbert-Schmidt space L02(U,L2(Rd)) if ψL(U,L2(Rd)) and

    ψ2L02=i=1λiψei22<.

    Lemma 3. ([7]) Let ϕ:[0,T]L02(U,L2(Rd)) satisfy T0ϕ(s)2L02ds<. Then integral t0ϕ(s)dBH(s) is a well defined random variable which takes values in L2(Rd) and

    Et0ϕ(s)dBH(s)222Ht2H1t0ϕ(s)2L02ds,t[0,T]. (2.1)

    Let (Y,B(Y),π) be a σ-finite measurable space on (Ω,F,P) and {kt}t0 be a stationary Poisson point process on (Y,B(Y),π). Let η(dy,t) be a Poisson counting measure which is generated by {kt}t0. The Poisson martingale measure ˜η on (Y,B(Y),π) is given by

    ˜η(dy,t)=η(dy,t)π(dy)t.

    Let Sπ,2(Y,L2(Rd)) be the space of all measurable functions p defined on (Y,B(Y),π) with

    p2Sπ,2=Yp(x,u;y)22π(dy)<.

    We use notation Sπ,2F([0,T]×Y,L2(Rd)) to denote the space of all F×B(Y)-measurable processes q with

    q2Sπ,2F=ET0Yq(s,x,u;y)22π(dy)ds<.

    The integral

    t+0Yq(s,x,u;y)˜η(dy,ds),t[0,T]

    is well defined when qSπ,2F([0,T]×Y,L2(Rd)). Further, if qSπ,2F([0,T]×Y,L2(Rd)), then for all t[0,T], t+0Yq(s,u;y)˜η(dy,ds) belongs to L2(Ω;L2(Rd)) and

    Et+0Yq(s,u;y)˜η(dy,ds)22=Et0Yq(s,u;y)22π(dy)ds. (2.2)

    See [22,23] for details.

    In this section, we give the representation of mild solutions of problem (1.1). Under suitable assumptions, we obtain the existence and uniqueness of mild solutions by iterative technique.

    Let DT be the set of all random processes having the following properties:

    (1) u is Ft adapted for any t[0,T].

    (2) uLFt2([0,T],L2(Rd)).

    (3) u is right continuous with left limit for all xRd and a.s. ωΩ, that is,

    u(t,x)=limstu(s,x).

    Definition 1. We call uDT a mild solution of (1.1) if u satisfies the following formula a.s.

    u=Rβα(t)u0+t0Sβα(ts)f(s,u)ds+t0Sβα(ts)Avds+t0Sβα(ts)g(s)dBH(s)+t+0YSβα(ts)h(s,u;y)˜η(dy,ds) (3.1)

    or equivalently

    u(t,x)=RdΦ(t,xz)u0(z)dz+t0RdΨ(ts,xz)f(s,z,u(s,z))dzds+t0RdΨ(ts,xz)Av(z)dzds+t0RdΨ(ts,xz)g(s,z)dzdBH(s)+t+0YRdΨ(ts,xz)h(s,z,u(s,z);y)dz˜η(dy,ds). (3.2)

    Remark 2. In fact, a mild solution given by Definition 1 is a formal solution of problem (1.1). Furthermore, if u is a classical solution of problem (1.1), i.e., u is sufficiently smooth and u satisfies problem (1.1) for all (t,x)[0,)×Rd, then u must be a mild solution of problem (1.1).

    The assumptions are given as follows.

    (A1): There exists a positive function Λ(x)L1(Rd)L2(Rd) such that

    |f(t,x,u(t,x))|2+Y|h(t,x,u(t,x);y)|2π(dy)Λ(x)(1+|u(t,x)|).

    (A2): There exists a positive real number L>0 such that

    |f(t,x,˜u(t,x))f(t,x,˜˜u(t,x))|2+Y|h(t,x,˜u(t,x);y)h(t,x,˜˜u(t,x);y)|2π(dy)L|˜u(t,x)˜˜u(t,x)|2.

    (G): Function g:[0,)L02(U,L2(Rd)) and there exists a constant r>1/(2β1) such that

    T0g(s,)2rL02ds<C(T),T>0,

    where C(T) denotes the constant dependent on T.

    (H): Function hSπ,2F([0,)×Y,L2(Rd)).

    We show the regularity of stochastic integral terms before introduce the main result.

    Lemma 4. Let β(1/2,1),α(1,2) and H(1/2,1). If (A1), (G) and (H) hold, then for any uDT,

    t0RdΨ(ts,xz)g(s,z)dzdBH(s),t+0YRdΨ(ts,xz)h(s,z,u(s,z);y)dz˜η(dy,ds)

    are well defined stochastic processes in DT.

    Proof. We divide the proof into two parts.

    (i) First, we prove that t0RdΨ(ts,xz)g(s,z)dzdBH(s)LFt2([0,T],L2(Rd)). By assumption (G), inequality (2.1), the Young inequality, the Hölder inequality and Lemma 1, we get that

    Et0RdΨ(ts,xz)g(s,z)dzdBH(s)22=Et0Ψ(ts)g(s)dBH(s)222HT2H1t0Ψ(ts,)21g(s,)2L02ds2HT2H1+r(2β1)1r(T0g(s,)2rL02ds)1/rC(T,H,β,r).

    Next we show that t0RdΨ(ts,xz)g(s,z)dzdBH(s)C([0,T];L2(Rd)). For any ε>0, we consider

    E{t+ε0RdΨ(ts+ε,xz)g(s,z)dzdBH(s)t0RdΨ(ts,xz)g(s,z)dzdBH(s)22}2E{t0Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]g(s,z)dzdBH(s)22}+2E{t+εtRdΨ(ts+ε,xz)g(s,z)dzdBH(s)22}G1+G2.

    By inequality (2.1) and the Young inequality, for any sufficient small δ>0, which is independent on ε, we get that

    G1=2E{t0[Ψ(ts+ε)Ψ(ts)]g(s)dBH(s)22}4Ht2H1t0Ψ(ts+ε,)Ψ(ts,)21g(s,)2L02ds4Ht2H1tδ0Ψ(ts+ε,)Ψ(ts,)21g(s,)2L02ds+4Ht2H1ttδΨ(ts+ε,)Ψ(ts,)21g(s,)2L02dsG1,1+G1,2.

    One can check that

    G1,14Ht2H1+r(2β1)1r{tδ0g(s,)2rL02ds}1/r<+.

    Since Ψ(t,x)C((0,),L1(Rd)), it is uniformly continuous on [δ,T]. Hence G1,10asε0. On the other hand, it is obvious that

    G1,24Ht2H1+r(2β1)1r{ttδg(s,)2rL02ds}1/r0asδ0.

    By the arbitrariness of δ, we get that

    G10asε0.

    By similar computations, we obtain

    G24Hε2H1t+εtΨ(ts+ε,)21g(s,)2L02ds0asε0.

    Hence t0RdΨ(ts,xz)g(s,z)dzdBH(s) is a well defined stochastic process in DT.

    (ii) Denote Xtt+0YRdΨ(ts,xz)h(s,z,u(s,z);y)dz˜η(dy,ds). We state that Xt is a well defined stochastic process which is right continuous with left limit. Let

    Ht,x(s,y)=RdΨ(ts,xz)h(s,z,u(s,z);y)dz.

    We assert that Ht,x(s,y)Sπ,2F([0,T]×Y,L2). By assumption (H), the Young inequality and Lemma 1, we arrive at

    Et0YHt,x(s,y)22π(dy)dsEt0YΨ(ts)21h(s;y)22π(dy)ds<.

    The assertion is proved.

    We now claim that XtLFt2([0,T],L2(Rd)). Following identity (2.2), the Schwartz inequality, the Fubini Theorem and assumption (A1), we deduce that

    ERd[t+0YRdΨ(ts,xz)h(s,z,u(s,z);y)dz˜η(dy,ds)]2dx=ERdt0Y|RdΨ(ts,xz)h(s,z,u(s,z);y)dz|2π(dy)dsdxERdt0Y{RdΨ(ts,xz)dz}{RdΨ(ts,xz)[h(s,z,u(s,z);y)]2dz}π(dy)dsdxc(T,d)ERdt0[RdΨ(ts,xz)dx]Λ(z)[1+|u(s,z)|]dzdsc(T,d)T0{Λ1+Λ2E[u(t,)2]}dsc(T,d).

    We remain to discuss the regularity of Xt. Since Xt is defined as a right limit, it is right continuous on t. Furthermore, by the continuity theorem in [22, Theorem 6.9], there exists a square integrable Ft-adapted martingale ˆXt which is right continuous with left-hand limit and satisfies

    P{ˆXt=Xt}=1,t[0,T].

    Therefore, Xt is right continuous with left limit for all xRd and a.s. ωΩ in this sense. Hence we conclude that XtDT.

    We introduce our main result of this section as follows.

    Theorem 1. Let β(1/2,1),α(1,2) and H(1/2,1). If (A1), (A2), (G) and (H) hold, then problem (1.1) has a unique mild solution u, subject to u0L2(Ω,L2(Rd)) and vV.

    Proof. We use the iterative technique and the Gronwall inequality to obtain the existence and uniqueness of mild solution. There are three steps to prove the theorem.

    Step 1. The sequence of iterations.

    Let

    u1(t,x)=Rβα(t)u0=RdΦ(t,xz)u0(z)dz,un+1(t,x)=u1(t,x)+t0RdΨ(ts,xz)f(s,z,un(s,z))dzds+t0RdΨ(ts,xz)(Av)(z)dzds+t0RdΨ(ts,xz)g(s,z)dzdBH(s)+t+0YRdΨ(ts,xz)h(s,z,un(s,z);y)dz˜η(dy,ds).u1(t,x)+4i=1Ji(t,x),

    where n=1,2,. Next, we show unDT, for any nN. From Lemma 2, Rβα(t)u0C([0,T],L2(Rd)) and

    Rβα(t)u0C([0,T],L2(Rd))u02

    and further u1DT. Let unDTforsomen>1. We need to prove un+1DT. By Lemma 4, we get that J3(t,x),J4(t,x)DT. Consequently, we are left with J1(t,x) and J2(t,x). By the Schwartz inequality, the Fubini Theorem, (A1) and Lemma 1, we obtain

    ERd|t0RdΨ(ts,xz)f(s,z,un(s,z))dzds|2dxtERdt0|RdΨ(ts,xz)f(s,z,un(s,z))dz|2dsdxtERdt0{RdΨ(ts,xz)dz}{RdΨ(ts,xz)|f(s,z,un(s,z))|2dz}dsdxC(T,d)ERdt0[RdΨ(ts,xz)dx]Λ(z)(1+|un(s,z)|)dzdsC(T,d)ET0{Λ1+Λ2un(s,)2}dsC(T,d).

    Hence J1LFt2([0,T],L2(Rd)). On the other hand, for arbitrarily ε>0, we consider

    EJ1(t+ε,x)J1(t,x)222ERd|t0Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]f(s,z,un)dzds|2dx+2ERd|t+εtRdΨ(ts+ε,xz)f(s,z,un)dzds|2dxΨ1+Ψ2.

    By the Schwartz inequality, the Fubini Theorem, (A1) and Lemma 1 we get that for any sufficient small δ>0,

    Ψ1C(T,d)ERdt0Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]|f(s,z,un(s,z))|2dzdsdxC(T,d)ERdt0Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]Λ(z)(1+|un(s,z)|)dzdsdxC(T,d)ERdt0{Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]dx}|Λ(z)|2dzds+C(T,d)ERdt0{Rd[Ψ(ts+ε,xz)Ψ(ts,xz)]dx}(1+|un(s,z)|2)dzdsC(T,d)Et0Ψ(ts+ε,)Ψ(ts,)1dsC(T,d)E{tδ0+ttδ}Ψ(ts+ε,)Ψ(ts,)1ds.

    Similar to the discussion for G1, we get that Ψ10 as ε0 and

    Ψ2C(T)t+εt(ts+ε)β1dsC(T)εβ0asε0.

    Therefore, J1DT. Since control vV and operator AL(V;L2(Rd)), we deduce that

    ERd|t0RdΨ(ts,xz)(Av)(z)dzds|2dxC(T)t0Ψ(ts,)21Av22dsC(T),

    by the Schwartz inequality and the Young inequality. So it is easily to see that J2(t,x)DT. We conclude that un+1DT. By induction, we get {un}n=1DT.

    Step 2. The existence

    (i) We show that the sequence of iterations is a Cauchy sequence in DT.

    Let Wn(t)=ERd[un+1(t,x)un(t,x)]2dx, for all 0tT. Note that

    W1(t)2CERdt0RdΨ(ts,xz)[|f(s,z,u1(s,z))|2+|Yh(s,z,u1(s,z);y)π(dy)|2]dzdsdx+2CHT2H1t0Ψ(ts,)21[Av22+g(s,)2L02]dsC(T,d)ET0{Λ1+Λ2u1(s,)2}ds+C(T,H)t0[Av22+g(s,)2L02]dsC(T,H,d),

    by (A1) and (G). For all n2, we assert that

    0Wn(t)|C(T)|n1(n1)!W1(T), (3.3)

    where 0tT. Suppose that (3.3) is true for n, we consider the case of n+1.

    Wn+1(t)2ERd{t0RdΨ(ts,xz)|f(s,z,un+1(s,z))f(s,z,un(s,z))|dzds}2dx+2ERd{t+0YRdΨ(ts,xz)|h(s,z,un+1(s,z);y)h(s,z,un(s,z);y)|dzπ(dy)ds}2dxCEt0Rd{Ψts,1}|un+1(s,z)un(s,z)|2dzdsCt0ERd|un+1(s,z)un(s,z)|2dzdsCt0Wn(s)ds|C(T)|nn!W1(T).

    Hence the assertion holds. Furthermore, formula (3.3) implies that {un(t,)} converges uniformly in L2(Rd) for any 0tT and a.s. ωΩ. The limit is denoted by u, i.e.

    u(t,)=limnun(t,)inL2(Rd).

    In addition, uDT.

    (ii) u is a mild solution.

    By assumption (A2), we get

    ERd{un(s,z)u1(s,z)t0RdΨ(ts,xz)f(s,z,u(s,z))dzdst0RdΨ(ts,xz)(Av)(z)dzdst0RdΨ(ts,xz)g(s,z)dzdBH(s)t0YRdΨ(ts,xz)h(s,z,u(s,z);y)dzπ(dy)ds}2dxERd{t0RdΨ(ts,xz)|f(s,z,un1(s,z))f(s,z,u(s,z))|dzds+t0YRdΨ(ts,xz)|h(s,z,un1(s,z);y)h(s,z,u(s,z);y)|dzπ(dy)ds}2dxc(T,L,d)t0ERd[un1(s,z)u(s,z)]2dzds0asn.

    We obtain that u satisfies (3.2), a.s.. Therefore, u is a mild solution of problem (1.1).

    Step 3. The uniqueness

    Let ˉu,ˉˉu be two different mild solutions of (1.1) with the same initial state u0 and control v. Let

    ζ(t)=ERd[ˉu(t,x)ˉˉu(t,x)]2dx.

    First, we can check that supt[0,T]ζ(t)<. Second, by similar arguments to that for (3.3), we get that

    ζ(t)Ct0ζ(s)ds.

    Hence we conclude that ˉu(t,x)=ˉˉu(t,x), for all t[0,T], a.e. xRd and a.s. ωΩ by the Gronwall inequality.

    In this section, we consider the existence of optimal control for problem (1.1) with a non-convex cost function. From Theorem 1, there exists a solution map: vu, for any vV. We call (u,v) a solution pair. We consider the cost function J has the following abstract form

    J(v)=E{T0P(t,u;v)dt+Q(u(T))}. (4.1)

    We assume that

    (J1) P:[0,T]×L2(Rd)×V[0,+). Functional P is measurable in t[0,T]. P(t,;η) is continuous, a.e. t[0,T] and uniformly for ηV. P(t,ξ;) is continuous and bounded, for a.e. t[0,T] and all ξL2(Rd). Further there exists a positive function aL1([0,T]) and a positive constant b>0 such that

    P(t,ξ;η)a(t)+bξ22,ηV.

    (J2) Q:L2(Rd)[0,+). Q is lower semi-continuous on L2(Rd) and there exist b1[0,) and b2(0,) such that

    Q(ξ)b1+b2ξ22.

    Let the admissible set Vad be a compact subset of V. If there exists an element v0Vad such that

    J(v0)=infvVadJ(v),

    then v0 is called an optimal control of J.

    Theorem 2. Let the conditions in Theorem 1 and (J1), (J2) hold. Then problem (1.1) with cost function (4.1) has at least one optimal control in Vad.

    Proof. Let J0infvVadJ(v). Then J00 and there exists {vn} such that J(vn)J0 as n. By the compactness of Vad, there exists a convergent subsequence of {vn}, still denoted by {vn}. Suppose vn˜v in Vad.

    Let {(un,vn)}n=1 and (˜u,˜v) be solution pairs. Next we prove un˜u in LFt2([0,T];L2(Rd)). Since {un}n=1 and ˜u are mild solutions, by similar arguments to the proof of (3.3), we obtain

    Eun(t,)˜u(t,)22C(T)T0E{un(s,)˜u(s,)22}ds+C(T)T0EΨ(ts,)21A(vn˜v)22ds.

    From the Gronwall inequality that

    Eun(t,)˜u(t,)22C(T)t0T0EΨ(τs,,)21A(vn˜v)22dsdτC(T,β)T0EA(vn˜v)22ds0 (4.2)

    as n, uniformly in t[0,T]. We arrive at un˜uasn in LFt2([0,T];L2(Rd)). Hence {un} is bounded in LFt2([0,T];L2(Rd)), i.e., there exists M>0 such that

    un2LFt2=E[supt[0,T]un(t,)22]<M,nN. (4.3)

    Especially, we get

    E[un(T,)22]<M,nN. (4.4)

    We show that J is lower semi-continuous. By assumption (J1) and (4.3), we get that

    ET0P(t,un;vn)dtT0a(t)dt+bET0un(t,)22dt<.

    Consider

    ET0P(t,˜u;˜v)dtET0[P(t,˜u;˜v)P(t,˜u;vn)]dt+ET0[P(t,˜u;vn)P(t,un;vn)]dt+ET0P(t,un;vn)dt. (4.5)

    Since P(t,ξ;) is continuous, by (J1) and the Lebesgue Dominated Convergence Theorem we obtain

    ET0[P(t,˜u;˜v)P(t,˜u;vn)]dt0asn. (4.6)

    As P(t,;η) is continuous, a.e. t[0,T] and uniformly in ηV,

    ET0[P(t,˜u;vn)P(t,un;vn)]dt0asn. (4.7)

    By (4.5)–(4.7), letting n we get

    ET0P(t,˜u;˜v)dtlim_nET0P(t,un;vn)dt. (4.8)

    On the other hand, by (4.2) and the lower semi-continuity of Q we get that

    Q(˜u(T))lim_nQ(un(T)),a.s.

    and by (J2)

    Q(un(T))b1+b2un(T)22,a.s.

    and further by (4.4) we have

    EQ(un(T))<M.

    Hence by the Fatou Lemma and (4.2), we conclude that

    EQ(˜u(T))lim_nEQ(un(T)). (4.9)

    It is clearly that J is lower semi-continuous by (4.8) and (4.9). Therefore, there exists v0Vad such that

    J(v0)=J0=infvVadJ(v).

    We get that v0 is an optimal control of problem (1.1) with cost function (4.1).

    In this section, two examples are given to illustrate the results. One is a simple example in 1-dimensional space, another is an abstract example in R3.

    Example 1. Consider the following fractional stochastic control system

    {Dc0D0.75tu+(Δ)α2u=u2+x2+t2dBH(t)dt,t[0,1],x[0,1],u=0,xR[0,1],u0=ex,x[0,1], (5.1)

    where α(1,2) and H(1/2,1).

    The order of Caputo fractional derivative is β=0.75 and t[0,1], then g=t2 satisfies assumption (G) for r(2,+). Functions f=u2 and h0 satisfy assumptions (A1), (A2) and (H). Hence, by Theorem 1, fractional stochastic control system (5.1) has an unique mild solution. Let Vad=[0,1], Av=v2, v=x and

    J=ET0u22dt. (5.2)

    It implies that P=u22 and Q is null operator. One can check that operators P and Q satisfy assumptions (J1) and (J2). Therefore, by Theorem 2, system (5.1) with cost function (5.2) has at least one optimal control.

    Example 2. We consider the following fractional stochastic control system

    {Dc0D0.6tu+(Δ)0.8u=a1u+Av+a2tdB0.75(t)dt+Ya3uy˜η(dy),t(0,1],xR3,u0=e|x|,xR3, (5.3)

    where a1,a2 and a3 are positive constants.

    Note that β=0.6,α=1.6 and H=0.75. Then g=a2t satisfies (G) for r(5,+). Further assume that

    Yy2˜η(dy)<.

    Thus f=a1u and h=a3uy satisfy (A1), (A2) and (H). Therefore, we deduce that there exists an unique mild solution of system (5.3). Let

    J(v)=E{T0eλtuθvθdt+a4uθ(T)}, (5.4)

    where λ(0,+) and θ(0,1). So P=eλtuθvθ and Q=a4uθ(T). It is easily to see that function J satisfies (J1) and (J2). By Theorem 2, there exists an optimal control of system (5.3) with cost function (5.4).

    In this paper, by iterative technique and energy estimates, we obtain the existence and uniqueness of mild solution to problem (1.1) in suitable framework and under Lipschitz type conditions. We prove the existence of optimal control for problem (1.1) with a non-convex cost function by nonlinear analysis method. At last, two examples are given to demonstrate applications of the theorems. It is well known that a weak solution is also a mild solution. However, the converse is not true. We will keep on discussing the properties of weak solutions to fully nonlocal stochastic problems.

    The authors are supported by the National Natural Science Foundation of China (Grant No. 11771107). We would like to thank referees for their valuable suggestions to improve the manuscript.

    The authors declare that there is no conflict of interest.



    [1] J. Sabatier, O. P. Agrawal, J. A. T. Machado, Advances in fractional calculus: Theoretical developments and applications in physics and engineering, Dordrecht: Springer, 2007.
    [2] B. B. Mandelbrot, J. W. Van Ness, Fractional Brownian motions, fractional noises and applications, SIAM Rev., 10 (1968), 422-437. doi: 10.1137/1010093
    [3] W. Dai, C. C. Heyde, Itô's formula with respect to fractional Brownian motion and its application, Int. J. Appl. Math. Stoch. Anal., 9 (1996), 439-448. doi: 10.1155/S104895339600038X
    [4] J. Kemppainen, J. Siljander, R. Zacher, Representation of solutions and large-time behavior for fully nonlocal diffusion equations, J. Differ. Equations, 263 (2017), 149-201. doi: 10.1016/j.jde.2017.02.030
    [5] L. Li, J. Liu, L. Wang, Cauchy problems for Keller-Segel type time-space fractional diffusion equation, J. Differ. Equations, 265 (2018), 1044-1096. doi: 10.1016/j.jde.2018.03.025
    [6] S. Tindel, C. Tudor, F. Viens, Stochastic evolution equations with fractional Brownian motion, Probab. Theory Relat. Fields, 127 (2003), 186-204. doi: 10.1007/s00440-003-0282-2
    [7] T. Caraballo, M. Garrido-Atienza, T. Taniguchi, The existence and exponential behavior of solutions to stochastic delay evolution equations with a fractional Brownian motion, Nonlinear Anal., 74 (2011), 3671-3684. doi: 10.1016/j.na.2011.02.047
    [8] L. Duc, M. Garrido-Atienza, A. Neuenkirch, B. Schmalfuß, Exponential stability of stochastic evolution equations driven by small fractional Brownian motion with Hurst parameter in (1/2, 1), J. Differ. Equations, 264 (2018), 1119-1145. doi: 10.1016/j.jde.2017.09.033
    [9] M. Tang, Q. Meng, Stochastic evolution equations of jump type with random coefficients: Existence, uniqueness and optimal control, Sci. China Ser. F: Inf. Sci., 60 (2017), 118202. doi: 10.1007/s11432-016-9107-1
    [10] A. Gu, D. Li, B. Wang, H. Yang, Regularity of random attractors for fractional stochastic reaction-diffusion equations on Rn, J. Differ. Equations, 264 (2018), 7094-7137. doi: 10.1016/j.jde.2018.02.011
    [11] L. Chen, Y. Hu, D. Nualart, Nonlinear stochastic time-fractional slow and fast diffusion equations on Rd, Stoch. Proc. Appl., 129 (2019), 5073-5112. doi: 10.1016/j.spa.2019.01.003
    [12] H. M. Ahmed, M. M. El-Borai, A. S. O. El Bab, M. E. Ramadan, Approximate controllability of noninstantaneous impulsive Hilfer fractional integrodifferential equations with fractional Brownian motion, Bound. Value Probl., 2020 (2020), 1–25. doi: 10.1186/s13661-019-01311-5
    [13] H. M. Ahmed, M. M. El-Borai, M. E. Ramadan, Boundary controllability of nonlocal Hilfer fractional stochastic differential systems with fractional Brownian motion and Poisson jumps, Adv. Differ. Equations, 2019 (2019), 1–23. doi: 10.1186/s13662-018-1939-6
    [14] H. M. Ahmed, Approximate controllability via resolvent operators of Sobolev-type fractional stochastic integrodifferential equations with fractional Brownian motion and Poisson jumps, Bull. Iran. Math. Soc., 45 (2019), 1045-1059. doi: 10.1007/s41980-018-0183-x
    [15] V. Barbu, M. Röckner, D. Zhang, Optimal control of nonlinear stochastic differential equations on Hilbert spaces, SIAM J. Control Optim., 58 (2020), 2383-2410. doi: 10.1137/19M1307615
    [16] G. Guatteri, F. Masiero, On the existence of optimal controls for SPDEs with boundary noise and boundary control, SIAM J. Control Optim., 51 (2013), 1909-1939. doi: 10.1137/110855855
    [17] P. Benner, C. Trautwein, A linear quadratic control problem for the stochastic heat equation driven by Q-Wiener processes, J. Math. Anal. Appl., 457 (2018), 776-802. doi: 10.1016/j.jmaa.2017.08.052
    [18] M. Fuhrman, Y. Hu, G. Tessitore, Stochastic maximum principle for optimal control of partial differential equations driven by white noise, Stoch. PDE: Anal. Comput., 6 (2018), 255-285. doi: 10.1007/s40072-017-0108-3
    [19] N. Durga, P. Muthukumar, Optimal control of fractional neutral stochastic differential equations with deviated argument governed by Poisson jumps and infinite delay, Optim. Control Appl. Methods, 40 (2019), 880-899. doi: 10.1002/oca.2515
    [20] G. M. Bisci, V. D. Rădulescu, R. Servadei, Variational methods for nonlocal fractional problems, Cambridge: Cambridge University Press, 2016.
    [21] D. Nualart, Stochastic calculus with respect to fractional Brownian motion, Ann. Fac. Sci. Toulouse Math., 15 (2006), 63-78. doi: 10.5802/afst.1113
    [22] N. Ikeda, S. Watanabe, Stochastic differential equations and diffusion processes, Amsterdam: North-Holland, 1981.
    [23] R. Situ, Theory of stochastic differential equations with jumps and applications: Mathematical and analytical techniques with applications to engineering, New York: Springer, 2005.
  • This article has been cited by:

    1. Dimplekumar Chalishajar, K. Ravikumar, K. Ramkumar, A. Anguraj, Null controllability of Hilfer fractional stochastic differential equations with nonlocal conditions, 2022, 0, 2155-3289, 0, 10.3934/naco.2022029
  • Reader Comments
  • © 2021 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(2315) PDF downloads(117) Cited by(1)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog