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

Finite element approximation of time fractional optimal control problem with integral state constraint

  • In this paper we investigate the finite element approximation of time fractional optimal control problem with integral state constraint. A space-time finite element scheme for the control problem is developed with piecewise constant time discretization and piecewise linear spatial discretization for the state equation. A priori error estimate for the space-time discrete scheme is derived. Projected gradient algorithm is used to solve the discrete optimal control problem. Numerical experiments are carried out to illustrate the theoretical findings.

    Citation: Jie Liu, Zhaojie Zhou. Finite element approximation of time fractional optimal control problem with integral state constraint[J]. AIMS Mathematics, 2021, 6(1): 979-997. doi: 10.3934/math.2021059

    Related Papers:

    [1] Yuelong Tang . Error estimates of mixed finite elements combined with Crank-Nicolson scheme for parabolic control problems. AIMS Mathematics, 2023, 8(5): 12506-12519. doi: 10.3934/math.2023628
    [2] Zuliang Lu, Ruixiang Xu, Chunjuan Hou, Lu Xing . A priori error estimates of finite volume element method for bilinear parabolic optimal control problem. AIMS Mathematics, 2023, 8(8): 19374-19390. doi: 10.3934/math.2023988
    [3] Tiantian Zhang, Wenwen Xu, Xindong Li, Yan Wang . Multipoint flux mixed finite element method for parabolic optimal control problems. AIMS Mathematics, 2022, 7(9): 17461-17474. doi: 10.3934/math.2022962
    [4] Zuliang Lu, Xiankui Wu, Fei Cai, Fei Huang, Shang Liu, Yin Yang . Error estimates in $ L^2 $ and $ L^\infty $ norms of finite volume method for the bilinear elliptic optimal control problem. AIMS Mathematics, 2021, 6(8): 8585-8599. doi: 10.3934/math.2021498
    [5] Chunjuan Hou, Zuliang Lu, Xuejiao Chen, Fei Huang . Error estimates of variational discretization for semilinear parabolic optimal control problems. AIMS Mathematics, 2021, 6(1): 772-793. doi: 10.3934/math.2021047
    [6] Changling Xu, Hongbo Chen . A two-grid $ P_0^2 $-$ P_1 $ mixed finite element scheme for semilinear elliptic optimal control problems. AIMS Mathematics, 2022, 7(4): 6153-6172. doi: 10.3934/math.2022342
    [7] Jun Moon . The Pontryagin type maximum principle for Caputo fractional optimal control problems with terminal and running state constraints. AIMS Mathematics, 2025, 10(1): 884-920. doi: 10.3934/math.2025042
    [8] Zuliang Lu, Fei Cai, Ruixiang Xu, Chunjuan Hou, Xiankui Wu, Yin Yang . A posteriori error estimates of hp spectral element method for parabolic optimal control problems. AIMS Mathematics, 2022, 7(4): 5220-5240. doi: 10.3934/math.2022291
    [9] Zuliang Lu, Fei Cai, Ruixiang Xu, Lu Xing . Convergence and proposed optimality of adaptive finite element methods for nonlinear optimal control problems. AIMS Mathematics, 2022, 7(11): 19664-19695. doi: 10.3934/math.20221079
    [10] Xingyang Ye, Chuanju Xu . A posteriori error estimates of spectral method for the fractional optimal control problems with non-homogeneous initial conditions. AIMS Mathematics, 2021, 6(11): 12028-12050. doi: 10.3934/math.2021697
  • In this paper we investigate the finite element approximation of time fractional optimal control problem with integral state constraint. A space-time finite element scheme for the control problem is developed with piecewise constant time discretization and piecewise linear spatial discretization for the state equation. A priori error estimate for the space-time discrete scheme is derived. Projected gradient algorithm is used to solve the discrete optimal control problem. Numerical experiments are carried out to illustrate the theoretical findings.


    This paper aims to develop a space-time finite element approximation of optimal control problem governed by time fractional diffusion equation and subjected to state constraint in integral form. Let Ω be a bounded domain of Rd(d=1,2,3) with sufficiently smooth boundary Ω. Set ΩT=Ω×(0,T). We consider the following optimal control problem governed by time-fractional diffusion equations:

    min(y,u)K×L2(ΩT)J(y,u):=12ΩT(y(x,t)yd(x,t))2dxdt+γ2ΩTu2(x,t)dxdt. (1.1)

    subject to

    {0Dαty(x,t)Δy(x,t)=f(x,t)+u(x,t),(x,t)ΩT,y(x,t)=0,(x,t)ΓT,y(x,0)=0,xΩ (1.2)

    and

    K={vL1(ΩT);ΩTvdxdtδ}. (1.3)

    Here yd is the desired state. γ>0 is the regularization constant. fL2(ΩT) is a given function.

    The research of control problem governed by differential equation forms a hot topic in the past decades. Lots of literatures are devoted to developing theoretical analysis or numerical methods([1,2,3,4,5,6,7]). As we know, fractional calculus has a very long history and was applied into many fields of science and engineering. For instance, the fractional PDEs arise in many engineering applications such as anomalous diffusion on fractals and fractional random walk([8,9]), and contaminant transport in groundwater flow([10,11]). Over the past decades lots of numerical methods and algorithms are developed to discretize fractional derivative and solve the fractional differential equations, including finite difference methods [12,13], finite element methods [14,15,16,17], spectral methods [18,19] and so on.

    In recent years optimal control problem governed by fractional PDEs has attracted the researcher's attention with the rapid development of numerical methods for fractional PDEs. For optimal control problem governed by time fractional PDEs we refer to [20,21,22,23,24] for finite element approximation, [25,28] for spectral Galerkin discretization. For optimal control problem governed by space fractional PDEs we refer to [26] for spectral Galerkin discretization, and [27] for finite element approximation. As far as we know, the researches on optimal control problem governed by fractional PDEs mainly focus on control constrained problem. The study with respect to state constrained fractional optimal control problem is very limited. In [28] spectral Galerkin approximation of time fractional optimal control problem with integral state constraint was discussed. In [29] an optimal control problem governed by fractional elliptic equation with pointwise state constraint was investigated. In [30] the authors discussed finite element approximation of space fractional optimal control problem with integral state constraint.

    In the present paper we consider an optimal control problem governed by time fractional diffusion equation with state constraint in integral form. The first order optimality condition is derived and the corresponding regularity is discussed. A space-time finite element discrete scheme is built up based on piecewise constant discontinuous Galerkin discretization for temporal discretization and conforming linear finite elements for spatial discretization. A priori error estimate for state, adjoint state and control variables is derived. Numerical example is given to verify the theoretical findings.

    The structure of this paper is as follows. In Section 2, the preliminaries of fractional derivative and Sobolev space are introduced. In Section 3 the space-time finite element approximation of the control problem is discussed. In Section 4 a priori error estimates are proved. Numerical example is given to verify the theoretical results in Section 5.

    In this section we introduce some basic knowledge of the fractional calculus and the Sobolev space.

    For 0<α<1, the left and right Riemann-Liouville fractional derivative are defined as follows

    0Dαtv=1Γ(1α)ddtt0v(s)(ts)αds,
    tDαTv=1Γ(1α)ddtTtv(s)(st)αds.

    Let Hδ(I) and Hδ0(I) denote the usual Sobolev space of order δ on I with norm Hδ(I) and semi-norm ||Hδ(I). According to [14], we have the following properties.

    Lemma 2.1. Assume that <a<b<. If vHα(a,b), then the following estimates hold

    aDα2tvL2(a,b)C|v|Hα2(a,b),tDα2bvL2(a,b)C|v|Hα2(a,b),aDα2tv,tDα2bv(a,b)C0|v|2Hα2(a,b).

    Further, if v,wHα2(a,b), then

    aDα2tv,tDα2bw(a,b)C|v|Hα2(a,b)|w|Hα2(a,b),aDαtv,wHα(a,b)=aDα2tv,tDα2bw(a,b)=v,tDαbwHα(a,b).

    We introduce the definition of space ˙Hs(Ω) following [31]. Let {λi}i=1 and {ϕi}i=1 denote eigenvalues and L2(Ω)-orthogonal eigenfunctions of Δ operator on domain Ω with zero Dirichlet boundary condition. For every vL2(Ω) we have the representation v=i=1aiϕi, where ai=(v,ϕi). For s0, let ˙Hs(Ω)L2(Ω) defined by

    ˙Hs(Ω):={v=i=1aiϕi:|v|s=(i=1a2iλsi)1/2<}.

    To define the weak formulation of state equation we also need to introduce the space

    V=Hα2(0,T;L2(Ω))L2(0,T;˙H1(Ω))

    and endow this space with the norm

    V=(||2Hα2(0,T;L2(Ω))+2L2(0,T;˙H1(Ω)))12.

    In this section we consider space-time finite element discretization of optimal control problem. To this end we begin with deriving the first order optimality condition and regularity of the solution.

    Theorem 3.1. Assume that (y,u) is the solution of control problem (1.1) and (1.2). Then there exist a real number ξ0 and the adjoint state p satisfying the following first order optimality system:

    {0DαtyΔy=f+u,(x,t)ΩT,y(x,t)=0,(x,t)ΓT,y(x,0)=0,xΩ, (3.1)
    {tDαTpΔp=yyd+ξ,(x,t)ΩT,p(x,t)=0,(x,t)ΓT,p(x,T)=0,xΩ, (3.2)
    (ξ,vy)ΩT0,vK (3.3)

    and

    γu+p=0. (3.4)

    Proof. According to [17], for given f+uL2(ΩT) the following weak formulation admits a unique solution yV

    0Dαty,vHα2(0,T;L2(Ω))+(y,v)ΩT=(f+u,v)ΩT,vV.

    We introduce the operator S mapping form L2(ΩT) to V, and the operator B mapping form V to L1(ΩT) (VL1(ΩT)). We define G:L2(ΩT)L1(ΩT) by G=SB. Then we can rewrite the optimal control problem as the following optimization problem

    minuL2(ΩT)ˆJ(u):=J(Su,u) (3.5)

    subject to

    GuK.

    In order to derive the first order optimality condition, we set H(u):=ΩTy(u)dxdtδ. Here y(u):=Su is the solution of the state equation. According to [32,33], there exists a real number ξ0 such that

    ξH(u)=0 (3.6)

    and

    Gu(u)(vu)=0,vL2(ΩT). (3.7)

    Here G(u,ξ)=ˆJ(u)+ξH(u) denotes the Lagrange functional with ξ being the Lagrange multiplier.

    By (3.6), we obtain

    0=ξH(u)=ξ(ΩTy(u)dxdtδ)=ξ(ΩT(y(u)v)dxdt)+ξ(ΩTvdxdtδ).

    Then we have

    (ξ,vy(u))=ξ(ΩTvdxdtδ)0.

    By a simple calculation we derive

    ˆJ(u)(vu)=limt0+ˆJ(u+t(vu))ˆJ(u)t=ΩT(y(u)yd)y(u)(vu)dxdt+γΩTu(vu)dxdt

    and

    H(u)(vu)=limt0+H(u+t(vu))H(u)t=ΩTy(u)(vu)dxdt.

    Thus we arrive at

    ΩT(y(u)yd)y(u)(vu)dxdt+γΩTu(vu)dxdt+ξΩTy(u)(vu)dxdt=0. (3.8)

    By the state equation, we deduce

    {0Dαty(u)(vu)Δy(u)(vu)=vu,in ΩT,y(u)(vu)=0,on ΓT,y(u)(vu)(x,0)=0,in Ω.

    To simplify (3.8), we introduce the adjoint state equation

    {tDαTpΔp=yyd+ξ,in ΩT,p(x,t)=0,on ΓT,p(x,T)=0,in Ω.

    By the integration by parts we deduce

    ΩT(yyd+ξ)y(u)(vu)dxdt=ΩT(tDαTpΔp)y(u)(vu)dxdt=ΩTp(0Dαty(u)(vu)Δy(u)(vu))dxdt=ΩTp(vu)dxdt.

    Combining this with (3.8), we have

    ΩTp(vu)dxdt+γΩTu(vu)dxdt=0.

    Thus we arrive at

    Gu(u)(vu)=ΩT(γu+p)(vu)dxdt=0.

    i.e.

    γu+p=0.

    Remark 3.2. According to (3.3), we can deduce that

    {ξ0,ΩTydxdt=δ,ξ=0,ΩTydxdt<δ.

    In the following we investigate the regularity of the solution to the optimal control problem. Assume that η is the solution of the state equation with right hand term g(x,t). Then according to [17] we have the following regularity estimates.

    Lemma 3.3. Assume that gL2(0,T;L2(Ω)),0<α<1. Then we have

    ηHα(0,T;L2(Ω))+|η|Hα2(0,T;˙H1(Ω))+ηL2(0,T;˙H2(Ω))CαgL2(0,T;L2(Ω)). (3.9)

    Furthermore, for gH1α(0,T;L2(Ω)),12<α<1, we have

    ηH1(0,T;L2(Ω))+|η|H1α2(0,T;˙H1(Ω))+ηL2(0,T;˙H2(Ω))Cα,TgH1α(0,T;L2(Ω)). (3.10)

    Remark 3.4. Note that ξ is a constant. Therefore the state and adjoint state have the same regularity to (3.9) for f,u,ydL2(0,T;L2(Ω)),0<α<1.

    Further we restrict that f,ydH1α(0,T;L2(Ω)),12<α<1. Note that γu=p and 12<α<1. Then by (3.9) we have uHα(0,T;L2(Ω))H1α(0,T;L2(Ω)). Combining with fH1α(0,T;L2(Ω)),12<α<1 leads to an improved regularity for the state variable y:

    yH1(0,T;L2(Ω))+|y|H1α2(0,T;˙H1(Ω))+yL2(0,T;˙H2(Ω))Cf+uH1α(0,T;L2(Ω)). (3.11)

    For ydH1α(0,T;L2(Ω)),12<α<1, above result further improves the regularity of the adjoint state variable p, i.e.,

    pH1(0,T;L2(Ω))+|p|H1α2(0,T;˙H1(Ω))+pL2(0,T;˙H2(Ω))Cyyd+ξH1α(0,T;L2(Ω)). (3.12)

    The weak formulation of the control problem can be characterized as:

    min(y,u)K×L2(ΩT)J(y,u) (3.13)

    subject to

    (0Dαty,v)ΩT+(y,v)ΩT=(f+u,v)ΩT,vL2(0,T;˙H1(Ω)). (3.14)

    Then the corresponding first order optimality system reads

    (0Dαty,v)ΩT+(y,v)ΩT=(f+u,v)ΩT,vL2(0,T;˙H1(Ω)), (3.15)
    (tDαTp,w)ΩT+(p,w)ΩT=(yyd,w)ΩT+ξ(1,w)ΩT,wL2(0,T;˙H1(Ω)), (3.16)
    ξ(1,vy)ΩT0,vK (3.17)

    and

    γu+p=0. (3.18)

    Let 0=t0<t1<...<tJ=T be a partition of [0, T] with τ=TJ. Set Ij:=(tj1,tj) for each 1jJ. Let Th be a quasi-uniform triangulation of Ω. We denote by h the maximum diameter of the elements in Th and define

    Mh={vH1(Ω):v|EP1(E),ETh},˜Vh={vhL2(0,T;Mh):vh(x,)Mh,vh(,t)|IjP0,1jJ},Vh={vhL2(0,T;MhH10(Ω)):vh(x,)Mh,vh(,t)|IjP0,1jJ}.

    Set Uh=VhK. Then the space-time finite element approximation of control problem can be characterized as follows

    min(yh, uh)UhטVhJ(yh,uh) (3.19)

    subject to

    (0Dαtyh,vh)ΩT+(yh,vh)ΩT=(f+uh,vh)ΩT,vhVh. (3.20)

    In an analogous way to continuous case we can derive the discrete first order optimality system:

    (0Dαtyh,vh)ΩT+(yh,vh)ΩT=(f+uh,vh)ΩT,vhVh, (3.21)
    (tDαTph,wh)ΩT+(ph,wh)ΩT=(yhyd,wh)ΩT+ξh(1,wh)ΩT,whVh, (3.22)
    ξh(1,vhyh)ΩT0,vhUh (3.23)

    and

    γuh+ph=0. (3.24)

    Remark 3.5. According to (3.23), the discrete Lagrange multipliers satisfies

    {ξh0,ΩTyhdxdt=δ,ξh=0,ΩTyhdxdt<δ.

    In this section we derive a priori error estimates for the space-time finite element discretization of the optimal control problem. For this purpose we need to introduce the following auxiliary problems for every wL2(0,T;˙H1(Ω))

    (0Dαty(uh),w)ΩT+(y(uh),w)ΩT=(f+uh,w)ΩT, (4.1)
    (tDαTp(uh),w)ΩT+(p(uh),w)ΩT=(y(uh)yd,w)ΩT+ξh(1,w)ΩT, (4.2)
    (tDαTp(yh),w)ΩT+(p(yh),w)ΩT=(yhyd,w)ΩT+ξh(1,w)ΩT. (4.3)

    For clarity we assume that f,ydL2(0,T;L2(Ω)) in the following analysis of Lemmas 4.1–4.3 and Theorems 4.4–4.5.

    Lemma 4.1. Assume that y and yh are the solutions of (3.15) and (3.21), respectively. Then we have

    yyhL2(0,T;L2(Ω))C(h2+τα)+CuuhL2(0,T;L2(Ω)).

    Proof. Combining (3.15) and (4.1) we have

    (0Dαt(yy(uh)),w)ΩT+((yy(uh)),w)ΩT=(uuh,w)ΩT.

    By Lemma 3.3 and the embedding theorem we have

    yy(uh)L2(0,T;L2(Ω))CuuhL2(0,T;L2(Ω)). (4.4)

    By (3.21) and (4.1) we get

    (0Dαt(y(uh)yh),vh)ΩT+((y(uh)yh),vh)ΩT=0.

    Since yh is the finite element approximation of y(uh), according to [17] we have

    y(uh)yhL2(0,T;L2(Ω))C(h2+τα). (4.5)

    By the triangle inequality we obtain

    yyhL2(0,T;L2(Ω))yy(uh)L2(0,T;L2(Ω))+y(uh)yhL2(0,T;L2(Ω))C(h2+τα)+CuuhL2(0,T;L2(Ω)).

    Lemma 4.2. Assume that (y,p,u,ξ) and (yh,ph,uh,ξh) are the solutions of the optimality system (3.15)–(3.18) and the discrete counterpart, respectively. Then we have

    pphL2(0,T;L2(Ω))C(h2+τα)+C(|ξξh|+uuhL2(0,T;L2(Ω))).

    Proof. By (3.16) and (4.3) we have

    (tDαT(pp(yh)),w)ΩT+((pp(yh)),w)ΩT=(yyh,w)ΩT+(ξξh,w)ΩT.

    By Lemma 2.1 and w=pp(yh) we have

    C0|pp(yh)|2Hα2(0,T;L2(Ω))+(pp(yh))2L2(0,T;L2(Ω))(yyhL2(0,T;L2(Ω))+|ξξh|)pp(yh)L2(0,T;L2(Ω)).

    This implies

    pp(yh)L2(0,T;L2(Ω))C(yyhL2(0,T;L2(Ω))+|ξξh|).

    Combining (3.22) and (4.3) we obtain

    (tDαT(p(yh)ph),wh)ΩT+((p(yh)ph),wh)ΩT=0.

    Note that ph is the finite element approximation of p(yh). Then we can derive by [17]

    p(yh)phL2(0,T;L2(Ω))C(h2+τα). (4.6)

    Using the triangle inequality and Lemma 4.1 we obtain

    pphL2(0,T;L2(Ω))C(h2+τα)+C|ξξh|+CuuhL2(0,T;L2(Ω)).

    Next we are going to derive the estimate of |ξξh| and uuhL2(0,T;L2(Ω)).

    Lemma 4.3. Assume that (y,p,u,ξ) and (yh,ph,uh,ξh) are the solutions of the optimality system (3.15)–(3.18) and the discrete counterpart, respectively. Then we have

    |ξξh|C(h2+τα)+CuuhL2(0,T;L2(Ω)).

    Proof. By (3.16) and (4.2) we have

    (tDαT(pp(uh)),w)ΩT+((pp(uh)),w)ΩT=(yy(uh),w)ΩT+(ξξh,w)ΩT. (4.7)

    Choosing w=ψV satisfy ψVC and 1|ΩT|ΩTψdxdt=1 and combining Lemma 3.3 we get

    |ξξh|C|(tDαT(pp(uh)),ψ)ΩT|+C|((pp(uh)),ψ)ΩT|+C|(yy(uh),ψ)ΩT|C|pp(uh)|Hα2(0,T;L2(Ω))+C(pp(uh))L2(0,T;L2(Ω))+Cyy(uh)L2(0,T;L2(Ω)). (4.8)

    Setting P=1|ΩT|ΩT(pp(uh))dxdt and choosing w=pp(uh)Pψ in (4.7) leads to

    (tDαT(pp(uh)),pp(uh))ΩT+((pp(uh)),(pp(uh)))ΩT=(tDαT(pp(uh)),Pψ)ΩT+((pp(uh)),(Pψ))ΩT+(yy(uh),pp(uh)Pψ)ΩT+(ξξh,pp(uh)Pψ)ΩT.

    Since (ξξh,pp(uh)Pψ)ΩT=0, then we can derive

    (tDαT(pp(uh)),pp(uh))ΩT+((pp(uh)),(pp(uh)))ΩT=(tDαT(pp(uh)),Pψ)+((pp(uh)),(Pψ))ΩT+(yy(uh),pp(uh)Pψ)ΩT.

    Using Lemma 2.1 we have

    C0|pp(uh)|2Hα2(0,T;L2(Ω))+(pp(uh))2L2(0,T;L2(Ω))C(|pp(uh)|Hα2(0,T;L2(Ω))|P|+(pp(uh))L2(0,T;L2(Ω))|P|+yy(uh)L2(0,T;L2(Ω))|P|)+yy(uh)L2(0,T;L2(Ω))pp(uh)L2(0,T;L2(Ω)).

    Note that pp(uh)L2(0,T;L2(Ω))C(pp(uh))L2(0,T;L2(Ω)). Further we have by Young inequality

    C0|pp(uh)|2Hα2(0,T;L2(Ω))+(pp(uh))2L2(0,T;L2(Ω))C(|pp(uh)|Hα2(0,T;L2(Ω))|P|+(pp(uh))L2(0,T;L2(Ω))|P|+yy(uh)L2(0,T;L2(Ω))(pp(uh))L2(0,T;L2(Ω))+yy(uh)L2(0,T;L2(Ω))|P|)12C0|pp(uh)|2Hα2(0,T;L2(Ω))+C|P|2+14(pp(uh))2L2(0,T;L2(Ω))+C|P|2+14(pp(uh))2L2(0,T;L2(Ω))+Cyy(uh)2L2(0,T;L2(Ω))+12yy(uh)2L2(0,T;L2(Ω))+12|P|212C0|pp(uh)|2Hα2(0,T;L2(Ω))+C|P|2+12(pp(uh))2L2(0,T;L2(Ω))+Cyy(uh)2L2(0,T;L2(Ω)).

    Then we derive

    C0|pp(uh)|2Hα2(0,T;L2(Ω))+(pp(uh))2L2(0,T;L2(Ω))C|P|2+Cyy(uh)2L2(0,T;L2(Ω)).

    By definition of P, (3.18), (3.24) and (4.5) we have

    |P|=|1|ΩT|ΩT(pp(uh))dxdt|CpphL2(0,T;L2(Ω))+Cphp(uh)L2(0,T;L2(Ω))CuuhL2(0,T;L2(Ω))+Cphp(yh)L2(0,T;L2(Ω))+Cp(yh)p(uh)L2(0,T;L2(Ω)).

    Combining (4.2) and (4.3) we get

    (tDαT(p(yh)p(uh)),w)ΩT+((p(yh)p(uh)),w)ΩT=(yhy(uh),w)ΩT.

    By Lemma 2.1 and w=p(yh)p(uh) we have

    p(yh)p(uh)L2(0,T;L2(Ω))Cyhy(uh)L2(0,T;L2(Ω)).

    Combined with the above inequality, we get

    |P|CuuhL2(0,T;L2(Ω))+Cphp(yh)L2(0,T;L2(Ω))+Cy(uh)yhL2(0,T;L2(Ω))C(h2+τα)+CuuhL2(0,T;L2(Ω)).

    Then we derive

    C0|pp(uh)|Hα2(0,T;L2(Ω))+(pp(uh))L2(0,T;L2(Ω))C|P|+Cyy(uh)L2(0,T;L2(Ω))C(h2+τα)+CuuhL2(0,T;L2(Ω)).

    Inserting the above estimate into (4.8) we obtain

    |ξξh|C(h2+τα)+CuuhL2(0,T;L2(Ω)). (4.9)

    Finally we need to estimate uuhL2(0,T;L2(Ω)).

    Theorem 4.4. Assume that (y,p,u,ξ) and (yh,ph,uh,ξh) be the solutions of (3.15)–(3.18) and the discrete counterpart, respectively. Then the following estimate holds

    uuhL2(0,T;L2(Ω))C(h2+τα).

    Proof. By (3.18) and (3.24) we have

    γuuh2L2(0,T;L2(Ω))=(γuγuh,uuh)ΩT=(php,uuh)ΩT=(p(yh)p,uuh)ΩT(p(yh)ph,uuh)ΩT.

    Combining (3.15) and (4.1) and choosing v=p(yh)p leads to

    (0Dαt(yy(uh)),p(yh)p)ΩT+((yy(uh)),(p(yh)p))ΩT=(uuh,p(yh)p)ΩT.

    Using (3.16) and (4.3), and setting w=yy(uh) yields

    (tDαT(p(yh)p),yy(uh))ΩT+((p(yh)p),(yy(uh)))ΩT=(yhy,yy(uh))ΩT+(ξhξ,yy(uh))ΩT.

    By using Lemma 2.1 we derive

    (0Dαt(yy(uh)),p(yh)p)ΩT+((yy(uh)),(p(yh)p))ΩT=(yhy,yy(uh))ΩT+(ξhξ,yy(uh))ΩT.

    Using above inequalities we further derive

    γuuh2L2(0,T;L2(Ω))=(yhy,yy(uh))ΩT+(ξhξ,yy(uh))ΩT(p(yh)ph,uuh)ΩT=(yhy(uh),yy(uh))ΩT+(y(uh)y,yy(uh))ΩT+(ξhξ,yhy(uh))ΩT(ξhξ,yhy)ΩT(p(yh)ph,uuh)ΩT.

    Further we have

    γuuh2L2(0,T;L2(Ω))+yy(uh)2L2(0,T;L2(Ω))=(yhy(uh),yy(uh))ΩT+(ξξh,y(uh)yh)ΩT(ξξh,yyh)ΩT(p(yh)ph,uuh)ΩT.

    Indeed, we have

    {ξ(1,yyh)ΩT0,ΩTydxdt=δ,ξ(1,yyh)ΩT=0,ΩTydxdt<δ

    and

    {ξh(1,yyh)ΩT0,ΩTyhdxdt=δ,ξh(1,yyh)ΩT=0,ΩTyhdxdt<δ.

    Then we have (ξξh,yyh)ΩT0. By Lemma 4.3 and Young inequality, we deduce

    γuuh2L2(0,T;L2(Ω))+yy(uh)2L2(0,T;L2(Ω))C(ε)(yhy(uh)2L2(0,T;L2(Ω))+p(yh)ph2L2(0,T;L2(Ω))+(h2+τα)2)+ε(yy(uh)2L2(0,T;L2(Ω))+uuh2L2(0,T;L2(Ω))).

    Using (4.4)–(4.6) we conclude

    uuhL2(0,T;L2(Ω))C(h2+τα).

    Based on above estimates, we can obtain the following result.

    Theorem 4.5. Assume that (y,p,u,ξ) and (yh,ph,uh,ξh) are the solution of (3.15)–(3.18) and (3.21)–(3.24), respectively. Then the following estimate holds

    yyhL2(0,T;L2(Ω))+pphL2(0,T;L2(Ω))+|ξξh|C(h2+τα).

    Proof. Combining Lemma 4.1–4.4 leads to the theorem results.

    Remark 4.6. If f,ydH1α(0,T;L2(Ω)), according to [17], we get y(uh)yhL2(0,T;L2(Ω))C(h2+τ) and p(yh)phL2(0,T;L2(Ω))C(h2+τ). By analogy with the above error estimate derivation process, we get the following error estimates

    yyhL2(0,T;L2(Ω))+pphL2(0,T;L2(Ω))+|ξξh|C(h2+τ).

    In this section, we use the projected gradient algorithm ([5]) to effectively solve the optimal control problem.

    Set the objective gradient function dk=γukh+pkh. Then we have

    uk+1h=ukhρdk,ρ>0.

    Here

    (0Dαtykh,vh)ΩT+(ykh,vh)ΩT=(f+ukh,vh)ΩT,vhVh

    and

    (tDαTpkh,wh)ΩT+(pkh,wh)ΩT=(ykhyd,wh)ΩT+ξkh(1,wh)ΩT,whVh.

    We set pkh=ˆpkh+ξkhψh with

    (tDαTˆpkh,wh)ΩT+(ˆpkh,wh)ΩT=(ykhyd,wh)ΩT,whVh

    and

    (tDαTψh,wh)ΩT+(ψh,wh)ΩT=(1,wh)ΩT,whVh.

    Further, we define ˆyk+1h by

    (0Dαtˆyk+1h,vh)ΩT+(ˆyk+1h,vh)ΩT=(f+ˆuk+1h,vh)ΩT,vhVh.

    Here ˆuk+1h=ukhρ(γukh+ˆpkh). Note that

    (0Dαtyk+1h,vh)ΩT+(yk+1h,vh)ΩT=(f+uk+1h,vh)ΩT,vhVh.

    Then we have

    yk+1h=ˆyk+1hρξkhφh.

    Here φh satisfies

    (0Dαtφh,wh)ΩT+(φh,wh)ΩT=(ψh,wh)ΩT,whVh.

    In order to guarantee the state constraints, we choose ξkh as follows

    ξkh=1ρΩTφhdxdtmax{ΩTˆyk+1hdxdtδ,0}.

    It is easy to check that ΩTyk+1hdxdtδ. The projected gradient algorithm is described in detail.

    Algorithm 1:
    1: Given initial value u0h=0, and parameter ρ>0. Solve the following equation:
    (tDαTψh,wh)ΩT+(ψh,wh)ΩT=(1,wh)ΩT,whVh,(0Dαtφh,wh)ΩT+(φh,wh)ΩT=(ψh,wh)ΩT,whVh.

    2: Solve the following equation:
    (0Dαtykh,vh)ΩT+(ykh,vh)ΩT=(f+ukh,vh)ΩT,vhVh,(tDαTˆpkh,wh)ΩT+(ˆpkh,wh)ΩT=(ykhyd,wh)ΩT,whVh.

    3: Compute ˆuk+1h=ukhρ(γukh+ˆpkh) and solve the following equation:
    (0Dαtˆyk+1h,vh)ΩT+(ˆyk+1h,vh)ΩT=(f+ˆuk+1h,vh)ΩT,vhVh.
    4: Compute
    ξkh=1ρΩTφhdxdtmax{ΩTˆyk+1hdxdtδ,0}.
    and pkh=ˆpkh+ξkhψh.
    5: Compute uk+1h=ukhρ(γukh+pkh). and solve the following equation:
    (0Dαtyk+1h,vh)ΩT+(yk+1h,vh)ΩT=(f+uk+1h,vh)ΩT,vhVh.
    6: If the stop condition is reached, end the loop. Otherwise, update k=k+1 and go to step (2).

    Example 5.1. In this example, we consider the optimal control problem with Ω=[0,1],T=1. The exact solution are given by

    y=tssin(2πx),p=(1t)ssin(πx),u=(1t)ssin(πx).

    In this example we consider a uniform partition for space with h=1M. We first consider the case with smooth solution. Tables 14 present the errors and convergence orders of state, control and Lagrange multiplier for α=13 and 23 with s=2. Tables 5 and 6 present the errors and convergence orders of state, control and Lagrange multiplier for α=0.4 with s=0.51. We can observe that the convergence rates with respect to time and space variable are in agreement with the theoretical findings predicted in Remark 4.6, i.e., first order convergence in time and second order convergence in space.

    Table 1.  L2(0,T;L2(Ω)) errors and convergence rates for space variable with α=1/3 and s=2.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06827755 0.07066092 0.45135857
    24 25 0.01675417 2.0269 0.01712133 2.0451 0.12283095 1.8776
    25 27 0.00417107 2.0060 0.00419205 2.0301 0.03229044 1.9275
    26 29 0.00104207 2.0010 0.00103673 2.0156 0.00827863 1.9636
    27 211 2.6052e-04 2.0000 2.5779e-04 2.0077 0.00209525 1.9823

     | Show Table
    DownLoad: CSV
    Table 2.  L2(0,T;L2(Ω)) errors and convergence rates for time variable with α=1/3 and s=2.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06827755 - 0.07066092 - 0.45135857 -
    24 24 0.02986199 1.1931 0.03384772 1.0619 0.25331048 0.8334
    25 25 0.01384412 1.1090 0.01663089 1.0252 0.13259718 0.9339
    26 26 0.00665006 1.0578 0.00826286 1.0092 0.06748221 0.9745
    27 27 0.00325723 1.0297 0.00412156 1.0035 0.03399226 0.9893

     | Show Table
    DownLoad: CSV
    Table 3.  L2(0,T;L2(Ω)) errors and convergence rates for space variable with α=2/3 and s=2.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06768648 - 0.06859494 - 0.44263663 -
    24 25 0.01662731 2.0253 0.01660560 2.0464 0.12156339 1.8644
    25 27 0.00414497 2.0041 0.00406390 2.0307 0.03221745 1.9158
    26 29 0.00103661 1.9995 0.00100463 2.0162 0.00831098 1.9548
    27 211 2.5934e-04 1.9990 2.4975e-04 2.0081 0.00211248 1.9761

     | Show Table
    DownLoad: CSV
    Table 4.  L2(0,T;L2(Ω)) errors and convergence rates for time variable with α=2/3 and s=2.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06768648 - 0.06859494 - 0.44263663 -
    24 24 0.02962587 1.1920 0.03279230 1.0647 0.25001564 0.8241
    25 25 0.01375176 1.1072 0.01610694 1.0257 0.13154320 0.9265
    26 26 0.00661364 1.0561 0.00800377 1.0089 0.06725091 0.9679
    27 27 0.00324272 1.0282 0.00399304 1.0032 0.03401301 0.9835

     | Show Table
    DownLoad: CSV
    Table 5.  L2(0,T;L2(Ω)) errors and convergence rates for space variable with α=0.4 and s=0.51.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06128791 - 0.07655843 - 0.58083469 -
    24 25 0.01672425 1.8737 0.02362219 1.6964 0.13448958 2.1106
    25 27 0.00447242 1.9028 0.00681702 1.7929 0.03242069 2.0525
    26 29 0.00118203 1.9198 0.00190351 1.8405 0.00795783 2.0265
    27 211 3.0990e-04 1.9314 5.2116e-04 1.8689 0.00197047 2.0138

     | Show Table
    DownLoad: CSV
    Table 6.  L2(0,T;L2(Ω)) errors and convergence rates for time variable with α=0.4 and s=0.51.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 23 0.06128791 - 0.07655843 - 0.58083469 -
    24 24 0.02666199 1.2008 0.04268519 0.8428 0.28928895 1.0056
    25 25 0.01288063 1.0496 0.02344623 0.8644 0.14138401 1.0329
    26 26 0.00654646 0.9764 0.01269053 0.8856 0.06895288 1.0359
    27 27 0.00339134 0.9489 0.00679131 0.9020 0.03375646 1.0304

     | Show Table
    DownLoad: CSV

    Secondly we consider the nonsmooth case about time variable with s=α0.49. In this case the right hand term f and yd with respect to time belong to L2(0,T). The errors and convergence rates of state, control and Lagrange multiplier are listed in Tables 7 and 8 with α=2/3. In this case we can observe that the convergence rate for space variable and the convergence rate for time variable approach to 2 and α, which are in agreement with the theoretical findings given in Theorem 4.5.

    Table 7.  L2(0,T;L2(Ω)) errors and convergence rates for space variable with α=2/3 and s=α0.49.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    23 10 0.05186062 0.02626828 0.08453145
    24 102 0.01142312 2.1827 0.00623473 2.0749 0.02266773 1.8988
    25 103 0.00263417 2.1165 0.00157102 1.9886 0.00597317 1.9241
    26 104 6.3914e-04 2.0431 4.0675e-04 1.9495 0.00158371 1.9152

     | Show Table
    DownLoad: CSV
    Table 8.  L2(0,T;L2(Ω)) errors and convergence rates for time variable with α=2/3 and s=α0.49.
    M J yyhL2(0,T;L2(Ω)) order uuhL2(0,T;L2(Ω)) order |ξξh| order
    27 27 0.00498051 0.00522381 0.00402921
    28 28 0.00314271 0.6643 0.00339663 0.6210 0.00246073 0.7114
    29 29 0.00200124 0.6511 0.00221703 0.6155 0.00156024 0.6573
    210 210 0.00128416 0.6401 0.00145172 0.6109 0.00101600 0.6189

     | Show Table
    DownLoad: CSV

    In this paper we discussed a space-time finite element approximation of time fractional optimal control problem with integral state constraint. A priori error estimate for the discrete scheme is derived. Numerical examples are presented to illustrate the theoretical findings.

    In our future work we are going to investigate the finite element approximation of time fractional optimal control problem with integral state constraint taking the form Ωy(x,t)δ,t[0,T].

    The researth was supported by National Natural Science Foundation of China (No. 11971276) and National Natural Science Foundation of Shandong Province (No. ZR2016JL004, ZR2017MA020).

    The authors declare there is no conflict of interests.



    [1] W. Liu, N. Yan, A posteriori error estimates for distributed convex optimal control problems, Adv. Comput. Math., 15 (2001), 285-309. doi: 10.1023/A:1014239012739
    [2] W. Liu, N. Yan, A posteriori error estimates for control problems governed by Stokes equations, SIAM J. Numer. Anal., 40 (2003), 1850-1869.
    [3] E. Casas, J. Raymond, Error estimates for the numerical approximation of Dirichlet boundary control for semilinear elliptic equations, SIAM J. Control Optim., 45 (2006), 1586-1611. doi: 10.1137/050626600
    [4] W. Gong, M. Hinze, Z. Zhou, Finite element method and a priori error estimates for Dirichlet boundary control problems governed by parabolic PDEs, J. Sci. Comput., 66 (2016), 941-967. doi: 10.1007/s10915-015-0051-2
    [5] W. Liu, D. Yang, L. Yuan, C. Ma, Finite element approximation of an optimal control problem with integral state constraint, SIAM J. Numer. Anal., 48 (2011), 1163-1185.
    [6] D. Yang, et al., State-dependent switching control of delayed switched systems with stable and unstable modes, Math. Method Appl. Sci., 41 (2018), 6968-6983. doi: 10.1002/mma.5209
    [7] D. Yang, X. Li, J. Qiu, Output tracking control of delayed switched systems via state-dependent switching and dynamic output feedback, Nonlinear Anal-Hybri., 32 (2019), 294-305. doi: 10.1016/j.nahs.2019.01.006
    [8] J. Bouchaud, A. Georges, Anomalous diffusion in disordered media: Statistical mechanisms, models and physical applications, Physics Reports, 195 (1990), 127-293. doi: 10.1016/0370-1573(90)90099-N
    [9] E. Barkai, R. Metzler, J. Klafter, Form continuous time random walks to the fractional Fokker-Planck equation, Physical Review E Statal Physics Plasmas Fluids & Related Interdiplinary Topics, 61 (2000), 132-138.
    [10] D. A. Benson, S. W. Wheatcraft, M. M. Meerschaeert, The fractional order governing equations of lévy motion, Water Resour. Res., 36 (2000), 1413-1423. doi: 10.1029/2000WR900032
    [11] M. M. Meerschaert, A. Sikorskii, Stochastic models for fractional calculus, 2012.
    [12] Y. N. Zhang, Z. Z. Sun, H. L. Liao, Finite difference methods for the time fractional diffusion equation on non-uniform meshes, J. Comput. Phys., 265 (2014), 195-210. doi: 10.1016/j.jcp.2014.02.008
    [13] H. L. Qiao, Z. G. Liu, A. J. Cheng, Two unconditionally stable difference schemes for time distributed-order differential equation based on Caputo-Fabrizio fractional derivative, Adv. Differ. Equ., 2020 (2020), 1-17. doi: 10.1186/s13662-019-2438-0
    [14] V. Ervin, J. Roop, Variational formulation for the stationary fractional advection dispersion equation, Numer. Meth. Part. D. E., 22 (2010), 558-576.
    [15] B. Jin, R. Lazarov, Z. Zhou, Two fully discrete schemes for fractional diffusion and diffusion-wave equations with nonsmooth data, SIAM J. Sci. Comput., 38 (2016), A146-A170. doi: 10.1137/140979563
    [16] H. Chen, H. Wang, Numerical simulation for conservative fractional diffusion equations by an expanded mixed formulation, J. Comput. Appl. Math., 296 (2016), 480-498. doi: 10.1016/j.cam.2015.09.022
    [17] B. Li, H. Luo, X. Xie, Analysis of a time-stepping scheme for time fractional diffusion problems with nonsmooth data, SIAM J. Numer. Anal., 57 (2019), 779-798. doi: 10.1137/18M118414X
    [18] Y. Lin, C. Xu, Finite difference/spectral approximations for the time-fractional diffusion equation, J. Comput. Phys., 225 (2007), 1533-1552. doi: 10.1016/j.jcp.2007.02.001
    [19] B. Li, H. Luo, X. Xie, A time-spectral algorithm for fractional wave problems, J. Sci. Comput., 7 (2018), 1164-1184.
    [20] Z. Zhou, W. Gong, Finite element approximation of optimal control problems governed by time fractional diffusion equation, Comput. Math. Appl., 71 (2016), 301-318. doi: 10.1016/j.camwa.2015.11.014
    [21] Z. Zhou, C. Zhang, Time-stepping discontinuous Galerkin approximation of optimal control problem governed by time fractional diffusion equation, Numer. Algorithms, 79 (2018), 437-455. doi: 10.1007/s11075-017-0445-3
    [22] M. Gunzburger, J. Wang, Error analysis of fully discrete finite element approximations to an optimal control problem governed by a time-fractional PDE, SIAM J. Control Optim., 57 (2019), 241-263. doi: 10.1137/17M1155636
    [23] B. Jin, B. Li, Z. Zhou, Pointwise-in-time error estimates for an optimal control problem with subdiffusion constraint, IMA J. Numer. Anal., 40 (2020), 377-404. doi: 10.1093/imanum/dry064
    [24] C. Zhang, H. Liu, Z. Zhou, A priori error analysis for time-stepping discontinuous Galerkin finite element approximation of time fractional optimal control problem, J. Sci. Comput., 80 (2019), 993-1018. doi: 10.1007/s10915-019-00964-9
    [25] X. Ye, C. Xu, Spectral optimization methods for the time fractional diffusion inverse problem, Numer. Math. Theory Me., 6 (2013), 499-519. doi: 10.4208/nmtma.2013.1207nm
    [26] L. Zhang, Z. Zhou, Spectral Galerkin approximation of optimal control problem governed by Riesz fractional differential equation, Appl. Numer. Math., 143 (2019), 247-262. doi: 10.1016/j.apnum.2019.04.003
    [27] Z. Zhou, Z. Tan, Finite element approximation of optimal control problem governed by space fractional equation, J. Sci. Comput., 78 (2019), 1840-1861. doi: 10.1007/s10915-018-0829-0
    [28] X. Ye, C. Xu, A spectral method for optimal control problem governed by the abnormal diffusion equation with integral constraint on the state, Chinese science: Mathematics, 7 (2016), 1053-1070.
    [29] H. Antil, D. Verma, M. Warma, Optimal control of fractional elliptic PDEs with state constraints and characterization of the dual of fractional order Sobolev spaces, 186 (2020), 1-23.
    [30] Z. Zhou, J. Song, Y. Chen, Finite element approximation of space fractional optimal control problem with integral state constraint, Numer. Math. Theory Me., 13 (2020), 1027-1049. doi: 10.4208/nmtma.OA-2019-0201
    [31] V. Thomée, Galerkin Finite Element Methods for Parabolic Problems, Springer-Verlag, 1984.
    [32] E. Cacas, Pontryagin's principle for state-constrained boundary control problems of semilinear parabolic equations, SIAM J. Control Optim., 35 (1997), 1297-1327. doi: 10.1137/S0363012995283637
    [33] E. Casas, Error estimates for the numerical approximation of semilinear elliptic control problems with finitely many state constraints, ESAIM. Contr. Optim. Ca., 8 (2002), 345-374. doi: 10.1051/cocv:2002049
  • This article has been cited by:

    1. Zuliang Lu, Xiankui Wu, Fei Cai, Fei Huang, Shang Liu, Yin Yang, Error estimates in $ L^2 $ and $ L^\infty $ norms of finite volume method for the bilinear elliptic optimal control problem, 2021, 6, 2473-6988, 8585, 10.3934/math.2021498
    2. Xiaoyong Xu, Linchen Xiong, Fengying Zhou, Solving fractional optimal control problems with inequality constraints by a new kind of Chebyshev wavelets method, 2021, 54, 18777503, 101412, 10.1016/j.jocs.2021.101412
    3. Anis Galimyanov, Tatyana Gorskaya, E. Vdovin, Calculation of fractional integrals using partial sums of Fourier series for structural mechanics problems, 2021, 274, 2267-1242, 03011, 10.1051/e3sconf/202127403011
    4. Juan Antonio Rojas-Quintero, Jorge Villalobos-Chin, Victor Santibanez, Optimal Control of Robotic Systems Using Finite Elements for Time Integration of Covariant Control Equations, 2021, 9, 2169-3536, 104980, 10.1109/ACCESS.2021.3099131
    5. Dossan Baigereyev, Nurlana Alimbekova, Abdumauvlen Berdyshev, Muratkan Madiyarov, Convergence Analysis of a Numerical Method for a Fractional Model of Fluid Flow in Fractured Porous Media, 2021, 9, 2227-7390, 2179, 10.3390/math9182179
    6. Junying Cao, Zhongqing Wang, Ziqiang Wang, A Uniform Accuracy High-Order Finite Difference and FEM for Optimal Problem Governed by Time-Fractional Diffusion Equation, 2022, 6, 2504-3110, 475, 10.3390/fractalfract6090475
    7. Mengya Su, Liuqing Xie, Zhiyue Zhang, Numerical Analysis of Fourier Finite Volume Element Method for Dirichlet Boundary Optimal Control Problems Governed by Elliptic PDEs on Complex Connected Domains, 2022, 10, 2227-7390, 4779, 10.3390/math10244779
    8. Jiaqi Zhang, Yin Yang, An efficient and accurate numerical method for the fractional optimal control problems with fractional Laplacian and state constraint, 2023, 39, 0749-159X, 4403, 10.1002/num.23056
    9. Nurlana Alimbekova, Abdumauvlen Berdyshev, Muratkan Madiyarov, Yerlan Yergaliyev, Finite Element Method for a Fractional-Order Filtration Equation with a Transient Filtration Law, 2024, 12, 2227-7390, 2519, 10.3390/math12162519
    10. Muhammad Bilal Hafeez, Marek Krawczuk, Fractional Spectral and Fractional Finite Element Methods: A Comprehensive Review and Future Prospects, 2024, 31, 1134-3060, 3443, 10.1007/s11831-024-10083-w
    11. Yuelong Tang, Error estimates of mixed finite elements combined with Crank-Nicolson scheme for parabolic control problems, 2023, 8, 2473-6988, 12506, 10.3934/math.2023628
    12. Lakhlifa Sadek, Said Ounamane, Bouchra Abouzaid, El Mostafa Sadek, The Galerkin Bell method to solve the fractional optimal control problems with inequality constraints, 2024, 77, 18777503, 102244, 10.1016/j.jocs.2024.102244
    13. Said Ounamane, Lakhlifa Sadek, Bouchra Abouzaid, El Mostafa Sadek, Extended truncated exponential method for solving tempered fractional variational problems, 2025, 1077-5463, 10.1177/10775463251320695
  • 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(3962) PDF downloads(200) Cited by(13)

Figures and Tables

Tables(8)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog