Abstract
This paper aims to retrieve the initial value for a non-local fractional Sobolev–Galpern problem. The given data are subject to noise by the discrete random model. We show that the solution to the problem is ill-posed in the sense of Hadamard. In this paper, we applied the Fourier truncation method to construct the regularized solution. We estimate the convergence between the solution and the regularized solution. In addition, the numerical example is also proposed to assess the efficiency of the theory.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In the last few decades, many scientists have demonstrated that fractional models describe natural phenomena in a more precise and systematic way than integer-order fraction models they are classic with regular time derivatives [1,2,3,4]. In this paper, we consider the integral boundary problem for the fractional Sobolev equation as follows:
where \(\rho > 0\), \(s \in (0,1)\), \(\Omega = (0,\pi ) \times (0,\pi ) \subset \mathbb {R}^{2}\) and \( \ell \in L^{2}(\Omega )\). The symbol \(\partial _{t}^{\beta }u\) is denoted by the Caputo derivative which is defined by
The symbol \(\Gamma (.)\) denotes the standard Gamma function, the two constants \( \theta _1,\theta _2 \ge 0 \) satisfying \( \theta _1^2+\theta _2^2>0\). In the main equation of problem (1.1), by taking \( \beta =s=1 \), the first equation becomes the classical pseudo-parabolic (or called Sobolev equation) in the following
Pseudo-parabolic equations have many applications in science and technology, especially in physical phenomena such as homogeneous fluids through a fissured rock, aggregation of populations, ...see, e.g., [17] and its references. To our knowledge, the results for Eq. (1.3) are quite abundant and huge. Most of the results consider and study the existence of solutions and the properties of solutions to the initial value problem.
For the reader’s convenience, we can state our problem (1.1) as follows. Let us assume that \(\ell \) is the given input data in some suitable space. We need to determine the initial value \(u(x_1, x_2, 0)\). Our study is based on the insights provided by the publication [34]. Let us look at the third condition of our problem, i.e., is repeated as follows
The above condition is also called a non-local in-time condition. Some observations and comments on what has just been said are as follows:
-
If \(\theta _1=1\) and \(\theta _2=0\), the condition (1.4) becomes the terminal condition
$$\begin{aligned} u(x_{1},x_{2},T)=\ell (x_{1},x_{2}) . \end{aligned}$$(1.5)Many PDEs with terminal value condition are called backward problem which is a famous problem in the application. We can refer the reader to some interesting papers on the terminal value problem of various PDEs, Kaltenbacher et al. [19, 20], Jia et al. [15], Liu et al. [43], Yamamoto, see [16, 23,24,25], Janno see [13, 14], Rundell et al. [29, 30], Amar Debbouche et al. [9, 12, 18, 22, 31, 37], Triet, see [38, 39], Chao Yang, [40], Au, see [41] and Thach, see [42], and references therein. The results of the pseudo-parabolic equation (1.3) with a terminal condition are still limited. The most interesting recent work of determining the initial function is of Tuan and Caraballo [6,7,8, 36, 44,45,46,47].
-
If \(\theta _1=0\) and \(\theta _2 =1\) then we obtain the condition
$$\begin{aligned} \int \limits _{0}^{T} u(x_{1},x_{2},t)~\hbox {d}t=\ell (x_{1},x_{2}) . \end{aligned}$$(1.6)The non-local condition (1.6) appears in some models in N. Dokuchaev’s works similar to his article [10].
-
The meaning of the appearance of the condition (1.4) with \(\theta _1>0\) and \(\theta _2>0\) has been thoroughly discussed in [34].
There are two main approaches to the problem of recovering the initial value for parabolic or pseudo-parabolic equation from observed data \(\ell \). When we observe experimentally the function \(\ell \), we will generate an approximation of this function in two cases. The first case is in the sense of the deterministic case, i.e., there exists a function \(\ell ^\epsilon \) of \(\ell \) such that the norm of the difference \(\ell ^\epsilon \) and \(\ell \) is smaller than \(\varepsilon \), where \(\varepsilon \) is the noise level. The second case is in the sense of random cases. This means that we obtain the observation data in the form
Solving methods for stochastic models are often more complex than for deterministic models. Continuation of main ideas from previous articles [32, 33], we continue to investigate the random noise problem for the model (1.1). From the complexity of this model, we require a lot of cumbersome calculations in this paper. In this study, we consider the input data in the analysis view as follows. Let
be discrete points in \(\Omega = (0,\pi ) \times (0,\pi )\). Here, the observed random data \(\big ( \widetilde{\ell }_{k l}, \widetilde{F}_{k l}(t)\big )\) are noise of the exact data \(\big (\ell (c_{k},d_{l}), F(c_{k},d_{l},t) \big )\) by satisfying the following models
where \(\mathcal Y_{kl}\) are mutually independent and identically distributed random variables and \(\mathcal X_{kl}(t)\) are independent one dimensional standard Brownian motions. The above random model is inspired by a similar model in [26, 27, 32, 33]. Due to the additional appearance of the integral in the non-local condition, the formulation of our solution becomes more complicated, which generates significant difficulties for us to handle when showing the ill-posedness as well as regularizing our problem. Our paper will investigate problem (1.1), and the main results of this work are as follows:
-
The proof for the ill-posedness of the solution and the estimate of the solution in \(L^{2}\big ( (0,\pi ) \times (0,\pi ) \big )\).
-
Proposing a regularized method and showing of convergence rate under a priori parameter choice rule.
This paper is organized as follows. Section 2 gives some preliminaries. In Sect. 3, we show the mild solution of problem (1.1), and an example describing the mismatch of the problem and the intercept of the source function in space \(L^{2}(\Omega )\) and in Sect. 4, we study the Fourier method to solve the problem (1.1) and show the rate of convergence. Finally, we show one numerical experiment to verify the theory results.
2 Preliminary
Let us collect some well-known properties of the negative Laplacian operator defined by
and recall the definition of the Hilbert scale space, which is a subset of \(L^2(\Omega )\). Since \(\mathcal {L}\) is a linear operator on \(\Omega = (0,\pi ) \times (0,\pi )\) with Dirichlet condition with the eigenvalues of \(\mathcal {L}\) are \(\lambda _{m,n} = m^{2} + n^{2}\), \(m,n \in \mathbb {N}\), and the corresponding eigenfunctions \( e _{m,n}\) satisfy
where \( e _{m,n}(x_{1},x_{2}) = e _{m}(x_{1}) e _{n}(x_{2}) = \dfrac{2}{\pi } \sin (m {x_1}) \sin (n{x_2}).\)
Definition 2.1
For any \(\nu > 0\), we define the fractional Hilbert scale space by
The space \( \mathcal {H}^{\nu }(\Omega ) \) is a Hilbert space equipped with the norm
In what follows, several properties of the Mittag–Leffler function are collected.
Definition 2.2
(see [21]) For \(0< \beta < 1\) and an arbitrary constant \( \alpha \in \mathbb {R}\), the Mittag–Leffler function is defined by
where \( \Gamma \) is the usual Gamma function.
Lemma 2.1
[35] For \( 0<\beta _1<\beta _2<1 \) and \(\beta \in [\beta _1,\beta _2],\) there exist positive constants \(\mathcal {M}_{1}\), \(\mathcal {M}_{2}\), and \( \mathcal {M}_{3}\), depending only on \( \beta \) such that
-
(i)
\(E_{\beta ,1}(-z)>0 \), for \( z>0. \)
-
(ii)
\( \dfrac{\mathcal {M}_{1}}{1+z} \le E_{\beta ,1}(-z)\le \dfrac{\mathcal {M}_{2}}{1+z},\) for \( z>0. \)
-
(iii)
\(E_{\beta ,\alpha }(-z) \le \dfrac{ \mathcal {M}_{3}}{1+z}\), for all \(z > 0, \alpha \in \mathbb {R}\).
Lemma 2.2
[28] For \(\lambda _{m,n} > 0\), \(1>\beta > 0\), and \(m,n \in \mathbb {N}\) are positive integers, we have:
Lemma 2.3
[34] Let \(\beta \in (0,1)\) and \(0< s < 1\), then we have
where \(\lesssim \) implies that “is less or equal up to a generic multiplicative constant C”.
Proof
The readers can refer to the document [34] for detailed proof. \(\square \)
Lemma 2.4
[35] For \(0< \beta < 1\), using Lemma 2.1, then we have
and putting \(\mathcal {M}:= \dfrac{\mathcal {M}_{2}}{\mathcal {M}_{1}}\Big (\dfrac{T}{t}\Big )^{\beta }\).
Lemma 2.5
[20] For \(0< \beta < 1\), \(\rho > 0\), \(E_{\beta ,1}\big (-\rho \big )\) is completely monotonic
Lemma 2.6
[20] For \(0< \beta < 1\), \(\rho > 0\), one has \(0 \le E_{\beta ,\beta }(-\rho ) \le \dfrac{1}{\Gamma (\beta )} \cdot \) Moreover \(E_{\beta ,\beta }\big (-\rho \big )\) is a monotonic decreasing function.
Lemma 2.7
(see Lemma 3.3 of [5]) Assume that \(m<i\), \(n<j\) and \(\Phi \in L^2(\Omega )\). Then,
Proof
Applying Lemma 3.5 of [11], page 145. This lemma can be proved. Therefore, we omit the detail here. \(\square \)
2.1 The Mild Solution of Problem (1.1)
Assume that problem (1.1) has a solution u of the following form
in which \(u_{m,n}(t) = \big <u(\cdot ,\cdot ,t), e _{m,n}\big >,~m,n \in \mathbb {N}^{*}\). Then, by applying the Laplace transform method, we obtain a formulation for the Fourier coefficients \(u_{m,n}(t)\) depending on \(\big <u_{0}, e _{m,n}\big>\) as follows:
From now on, for a shorter, we put \(a(\rho ,z) = {z^s~b(\rho ,z)}\) with \(b(\rho ,z) = \frac{1}{1+\rho z^{s}}\), \(z=\lambda _{m,n}\) then
Then, (2.15) can be written as follows:
This leads to
From the formula (2.18), using the condition non-local (1.4), through some transformations, we received
Substituting (2.19) into (2.18), we get
We have completed finding the mild solution to the problem (1.1).
2.2 Discretization Formula of the Solution
Here, we show a formula for the solution u with the discrete data \(\ell (c_{k},d_{l})\) and \(F(c_{k},d_{l},t)\) instead of the unknown data as in \(\ell _{m,n}\) and \(F_{m,n}(t)\). To do this, we need the following lemmas:
Lemma 2.8
Let m, n be positive integers such that \(m <i\) and \(n<j\). Setting
Then, the following property holds true
where \(\textbf{1}\) is the indicator function.
Proof
The lemma can be proved by using Lemma 3.5 in page 145 of [11]. \(\square \)
Theorem 2.1
Let positive integers \(\mathcal {N}_{tr},\mathcal {M}_{tr}\) such that \(\mathcal {N}_{tr} < i \) and \(\mathcal {M}_{tr} < j\). We assume \(\ell \in C^{1}(\overline{\Omega }) \cap L^{\infty }(0,T)\) and \(F \in C([0,T];C^{1}(\overline{\Omega }))\). Then, we have
where \(\mathcal {U}_{dis}\) is constructed based on discrete values of \(\ell \) and \(F(\cdot ,\cdot ,t)\) at \(i \times j\) grid points \((c_k,d_l) \)
the first residual sum \(\mathcal O^I_{res}\) is the difference \(\displaystyle \sum \nolimits _{m=1}^{ \mathcal M_{tr}} \sum \nolimits _{n=1}^{ \mathcal N_{tr}} u_{m,n}(t) e_{m,n} -\mathcal U_{dis}\)
and the second residual sum \(\mathcal O^{II}_{res}\) is the difference \(u- \displaystyle \sum \nolimits _{m=1}^{ \mathcal M_{tr}} \sum \nolimits _{n=1}^{ \mathcal N_{tr}} u_{m,n}(t) e_{m,n}\)
Here, \(\delta _{i,j,m,n}\) and \(\Theta _{i,j,m,n}(t)\) are the following differences
and
Proof
We replace \(\big <\ell (\cdot ,\cdot ), e _{m,n}\big>\) and \(\big <F(\cdot ,\cdot ,t), e _{m,n}\big>\) in the formula (2.20) by the data random in Lemma 2.9 and Lemma 2.10, this implies that the discretization formula of the solution u and we finish prove. \(\square \)
Lemma 2.9
Let \(\ell \) satisfies Theorem 2.1 and assume that \(m<i\), \(n<j\). Then, we get
Lemma 2.10
Let F satisfies Theorem 2.1 and assume that \(m<i\), \(n<j\), then we get
Proof
The demonstration of these two supplements is similar to in [32]. We omit here. \(\square \)
Remark 2.1
The explicit form of the mild solution (2.22) is more convenient (than (2.20)) for us to show the ill-posedness of our problem and construct an approximate solution later. We would like to emphasize that it is divided into two parts including the finite series and the bias between the two series, in which the first part \(\mathcal U_{dis}\) is expressed based on the value of \(\ell \) and \(F(\cdot ,\cdot ,t)\) at discrete point \((c_k,d_l)\) instead of the whole domain \(\Omega \). In this way, there appear several new terms in the residual term \(\mathcal O^{I}_{res}\) including \(\delta _{i,j,m,n}\) and \(\Theta _{i,j,m,n}(t)\) defined in (2.26) and (2.27). Fortunately, these couple of terms possess explicit forms like in Lemma 2.9.
3 The Non-well Posed of Problem (1.1) with Random Noise Data
Theorem 3.1
The problem (1.1) is ill-posed.
Proof
Let us design the exact data and the corresponding random noise data as follows:
and \(s=1\). Additionally, we design approximations of the exact data \(\ell \) and F at discrete points \((x_1,x_2)\) as follows:
Let u be the exact solution of Problem (1.1) with respect to the exact data (3.30)–(3.31). Then, it is obvious that is \(u \equiv 0\). Let \(\widetilde{u}^{i,j}\) be the solution of problem (1.1) with the data (3.32)–(3.33). Due to Lemma 2.9 with \(\mathcal N_{tr} = i - 1\) and \(\mathcal M_{tr} = j - 1\).
Next, we will show that the means \(\mathbb {E} \big \Vert \widetilde{\ell }^{i,j} - \ell \big \Vert _{ {L^2(\Omega )}}^{2}\) and \(\mathbb {E}\big \Vert \widetilde{F}^{i,j}(\cdot ,\cdot ,t) - F(\cdot ,\cdot ,t) \big \Vert ^{2}_{ {L^2(\Omega )}} \rightarrow 0\) but \(\mathbb {E}\big \Vert \widetilde{u}^{i,j}(\cdot ,\cdot ,t) - u(\cdot ,\cdot ,t) \big \Vert ^{2}_{ {L^2(\Omega )}} \nrightarrow 0\) as \(i,j \rightarrow \infty \) for any \(t \in [0,T]\), where \(\mathbb {E}\) denotes the exception.
\(\underline{{\textbf {Part 1:}}}\) We estimate that
First, we can find that
Use the properties of random variable \(\mathcal {Y}_{kl}\), we know \( = \mathbb {E}(\mathcal {Y}_{kl}\mathcal {Y}_{ij}) = \delta _{ki} \delta _{lj}\) for all k, l, i, j is the Kronecker delta, we have
From Lemma 2.8, it yields that
Therefore, we get
By using the similar techniques as shown in above, we have
Using the fact that
\(\underline{{\textbf {Part 2:}}}\) Estimate of \(\mathbb {E} \big \Vert \widetilde{u}^{i,j}(\cdot ,\cdot ,0) - u(\cdot ,\cdot ,0) \big \Vert _{L^{2}(\Omega )}^{2}\), we receive
We present the incorrectness of the test \(u(\cdot ,\cdot ,t)\) in case \(t=0\), from (3.41), one has
With the inequality \((a+b)^{2} \ge \dfrac{a^{2}}{2} - b^{2}\), one has
whereby
Now, we will estimate the error for \(\mathcal {L}_{1}\) and \(\mathcal {L}_{2}\) by using the formula (3.40); it gives
Next, thanks to Hölder’s inequality, we have
It gives
We know that
Hence, we can find that
It is easy to see that
Therefore, we get
On the other hand, it is clear that
Combining (3.43), and (3.44), and choosing \(s=1\), we conclude that
Thus, \(\mathbb {E} \big \Vert \widetilde{u}^{i,j}(\cdot ,\cdot ,0) - u(\cdot ,\cdot ,0) \big \Vert _{L^{2}(\Omega )} \rightarrow \infty \) as \(i,j \rightarrow \infty \). Therefore, the problem (1.1) is ill-posed in the Hadamard sense. \(\square \)
4 Regularized Solution and Convergent Rate
Theorem 4.1
Let \(\delta >2\), and let \( \mathcal {N}_{tr}, \mathcal {M}_{tr} \in \mathbb {Z}^{+}\) be such that \(\mathcal {N}_{tr}< i, \mathcal {M}_{tr} < j\). Assume that \(\ell \in \mathcal {H}^{\delta }(\Omega )\) and \(F \in L^{\infty }(0,T; {L^2(\Omega )})\), the problem (1.1) has a unique solution in \(C([0,T]; {L^2(\Omega )})\). The regularization solution is constructed as follows:
whereby
then we have the following estimate
Proof
Using the method of least square, we get
Remark 4.1
The \(\big \Vert \widetilde{U}_{\mathcal {N}_{tr},\mathcal {M}_{tr}}(\cdot ,\cdot ,t) - u(\cdot ,\cdot ,t) \big \Vert _{L^{2}(\Omega )}^{2}\) is of order
Next, we have to choose
in which \(\lfloor y \rfloor \) is greatest integer less than or equal y, then \(\big \Vert \widetilde{U}_{\mathcal {N}_{tr},\mathcal {M}_{tr}}(\cdot ,\cdot ,t) - u(\cdot ,\cdot ,t) \big \Vert _{L^{2}(\Omega )}^{2}\) is of order
Next, the probability of the Theorem (4.1) is done through some of the following steps:
Part 1: In this part, from (2.18)–(4.45), we have estimate
where \(\underbrace{\Psi _{m,n}^{i,j} = \overline{\ell }_{m,n}^{i,j} - \ell _{m,n}}_{{\textbf {Step 1:}}}\) and \(\underbrace{\Upsilon _{m,n}^{i,j} = \overline{F}_{m,n}^{i,j}(\tau ) - F_{m,n}(\tau )}_{{\textbf {Step 2:}}}\). After that, we conduct evaluation for \(\Psi _{m,n}^{i,j}\) và \(\Upsilon _{m,n}^{i,j}\), we get
\({\underline{{\textbf {Step 1:}}}}\) \(\mathbb {E} \big [\Psi _{m,n}^{i,j}\big ]^{2}\) can be bounded as follows:
By using familiar inequality \((a+b)^{2} \le 2a^{2}+2b^{2}\), for \(a,b \ge 0\), one has
Therefore, we notice that \(0 < \mathcal {B}_{min} \le \theta _{k,l} \le \mathcal {B}_{max}\), and \(\mathbb {E} \big (\mathcal {Y}_{kl} \mathcal {Y}_{pq}\big ) = \delta _{kp} \delta _{lq} \), addition
\(\dfrac{1}{ij} \sum \limits _{k=1}^{i} \sum \limits _{l=1}^{j} e _{m,n}^{2} (c_{k},d_{l}) = \pi ^{-2}\), we get
From the assumption \(\ell \in \mathcal {H}^{\nu }(\Omega )\), there exists a constant \(\mathcal {D}_{\nu }\), we have
From (4.52), this implies that
From (4.51) and the formula (2.26), we can see that
Estimate of \(\mathcal {A}_{1}\) as follows:
Combining (4.52) and (4.54), we conclude that
\(\underline{{\textbf {Step 2:}}}\) Evaluate \(\mathbb {E} \big [ \Upsilon _{m,n}^{i,j}(\tau ) \big ]^{2} \), we have also
Because of \(F \in L^{\infty }(0,T;L^{2}(\Omega ))\), this leads to
In here, \(\Theta _{i,j,m,n}\) is rated similarly (4.53)
We have known that \(\mathbb {E}\big ( {\mathcal {X}}_{kl}(\tau ) {\mathcal {X}}_{pq}(\tau ) \big ) = \delta _{kp}\delta _{ql} \tau \),, one has
From (4.49), we have
whereby
Estimate (4.59) will be shown right below; for ease and convenience, we divided the evaluation into two parts as follows:
Part 1 Estimate of \(\text {E}_{1}\), and due to \((a+b+c+d)^{2} \le 4 (a^{2}+b^{2}+c^{2}+d^{2})\), \(\forall a,b,c,d \ge 0\), we get that
Next, we continue to go to the evaluation \(\mathcal {B}^{m,n}_{i,j}\), \(\mathcal {C}_{i,j}^{m,n}(T)\) and \(\mathcal {D}^{m,n}_{i,j}(t)\) and \(\mathcal {E}^{m,n}_{i,j}(t)\). We evaluate the error (4.60) through four steps as follows.
\({\underline{Claim 1:}}\) First of all, from Lemmas 2.1 and 2.4, we have
This leads to
\({\underline{Claim 2:}}\) Next, we see that \(\mathcal {C}_{i,j}^{m,n}(T)\) can be bounded as follows:
Using the inequality \(b(\rho ,\lambda _{m,n}) \le 1\). Thus,
From (4.64), we have
\({\underline{Claim 3:}}\) Estimate of \(\big (\mathcal {D}_{i,j}^{m,n}\big )(t)\)
Hence, from Lemma 2.4, and combining (4.66) to (4.68), we get
\({\underline{Claim 4:}}\) Estimate of \(\mathcal {E}_{i,j}^{m,n}(t)\)
This implies that
and obviously
\({\underline{Claim 5:}}\) Similarly as in the evaluation (4.64), we can obtain
Combining (4.60) to (4.69), we conclude that
Part 2 Estimate of \(\text {E}_{2}\), assume that \(u \in C([0,T];\mathcal {H}^{\nu }(\Omega ))\), this means
It is easy to see that
Combining (4.70) to (4.71), we conclude that
The provision of this theorem is completed. \(\square \)
5 Simulation
We present numerical example, we choice \(\Omega = (0,\pi ) \times (0,\pi )\), \(T = 1\), \(\lambda _{m,n} = m^{2}+n^{2}\), \( e _{m,n}(x) = \dfrac{2}{\pi } \sin (mx) \sin (ny), s=1\), \(\theta _{1}=\theta _{2} = 0.45\), \(\beta =0.52\) are shown in this section. In this example, we test the problem as follows:
For the discretization, \((x_{i},y_{j})\) as discretion the space on \((0,\pi ) \times (0,\pi )\) for \(i=\overline{1,N_{x}+1}\), \(j=\overline{1,N_{y}+1}\) satisfied
whereby
The couple of \(\big (\ell _{\epsilon }(x_{1},x_{2}), F_{\epsilon }(x_{1},x_{2},t) \big )\), where we suppose that the measurements are described as follows:
whereby \( {\text {noise}_{1} } = \alpha \mathcal {N}(0,1)\). Besides that, \({\text {noise}_{2} } = \vartheta \Psi (t)\) with \(\Psi (t)\) is Brownian motion, with \( \vartheta , \alpha \) is a constant in (0, 1). The behavior of the input functions and their noise are shown in Fig. 1. Here, we use N(0, 1) as function numpy.random.randn(.) in Python software. With the ways to choose the function as in (5.74) and use the exact solution as in (2.20), we can easily find the u(x, y, t) function of the problem (5.73). In the same way, with the \(\ell _{\epsilon }(x_ {1}, x_{2}) \) functions and \( F_{\epsilon }(x_{1},x_{2},t)\) be given as in (5.75), we have the testing of \(u_ {\epsilon }(x_{1},x_{2},t)\), and during calculation, we also need to use the following integrals, see [21]
We present the absolute root mean square errors
When calculating these integrals, we also approximate how to calculate the integrals with Simpson integral as follows:
And with the way to choose \( N_{x} = N_{y} = 200\), we have matrix elements u represented as follows, similarly, we also represent \(u_{\epsilon }\) similar
The results of this section are presented in Table 1, and Figs. 1, 2, 3 and 4. In Figs. 1, 2, 3 and 4, we show an illustration of the input data of the problem and its approximation. In our calculation, we calculate for a specific case at \(t=0.5\), and present with full figures and table; however, at cases \(t=0.3\) and \(t=0.8\), the errors and drawings are also done similarly; therefore, we only present the calculation results in the number table and not presented with a drawing. Observe the drawings of input data and the exact solution and the regularized solution results, we commented that when the value of \(\epsilon \rightarrow 0\), the approximation will gradually converge the accuracy, this is not only true to the \(\ell \) function input data of non-local conditions but also true to the convergence of the regularization solution. From the observed data above, we can conclude that the proposed method is effective and stable.
6 Conclusion
In this article, we consider the Fourier truncation method for solving the backward problem (1.1). In this work, we have shown the non-well-posed of our problem (1.1), we present an example, applying the Fourier truncation method, and based on an a priori assumption for the exact solution, showing the error estimate between the exact solution and the regularized solution, having an additional numerical illustration will prove the correctness of our method.
References
Adiguzel, R.S., Aksoy, U., Karapinar, E., Erhan, I.M.: On the solutions of fractional differential equations via Geraghty type hybrid contractions. Appl. Comput. Math. 20, 313–333 (2021)
Adiguzel, R.S., Aksoy, U., Karapinar, E., Erhan, I.M.: Uniqueness of solution for higher-order nonlinear fractional differential equations with multi-point and integral boundary conditions. Revista de la Real Academia de Ciencias Exactas Físicas y Naturales Serie A Matemáticas 115, 1–16 (2021)
Baleanu, D., Amin, J., Hakimeh, M., Shahram, R.: A new study on the mathematical modelling of human liver with Caputo–Fabrizio fractional derivative. Chaos Solitons Fractals 134, 109705 (2020)
Baleanu, D., Sina, E., Shahram, R.: A hybrid Caputo fractional modeling for thermostat with hybrid boundary value conditions. Bound. Value Probl. 1, 1–16 (2020)
Binh, T.T., Can, H.N., Nam, D.H.Q., Thach, T.N.: Regularization of a two-dimensional strongly damped wave equation with statistical discrete data. Math. Methods Appl. Sci. 43, 4317–4335 (2020)
Caraballo, T., Guo, B., Tuan, N.H., Wang, R.: Asymptotically autonomous robustness of random attractors for a class of weakly dissipative stochastic wave equations on unbounded domains. Proc. R. Soc. Edinb. Sect. A 151, 1700–1730 (2021)
Caraballo, T., Tuan, N.H.: New results for convergence problem of fractional diffusion equations when order approach to \(1^-\). Differ. Integr. Equ. 36, 491–516 (2023)
Caraballo, T., Mchiri, L., Rhaima, M.: Ulam–Hyers–Rassias stability of neutral stochastic functional differential equations. Stochastics 94, 959–971 (2022)
Debbouche, A.: Time-partial differential equations: modeling and simulation. Math. Methods Appl. Sci. 44, 11767 (2021)
Dokuchaev, N.: On recovering parabolic diffusion from their time-averages. Calc. Var. Partial Differ. Equ. 58, 1–14 (2019)
Eubank, R.L.: Nonparametric Regression and Spline Smoothing. Statistics Textbooks and Monographs, vol. 157, 2nd edn. Marcel Dekker, Inc., New York (1999)
Guechi, S., Rajesh, D., Amar, D., Muslim, M.: Analysis and optimal control of \(\phi \)-Hilfer fractional semilinear equations involving nonlocal impulsive conditions. Symmetry 13, 2084 (2021)
Janno, J., Kinash, N.: Reconstruction of an order of derivative and a source term in a fractional diffusion equation from final measurements. Inverse Probl. 34, 025007 (2018)
Janno, J., Kasemets, K.: Uniqueness for an inverse problem for a semilinear time-fractional diffusion equation. Inverse Probl. Imaging 11, 125–149 (2017)
Jia, J., Peng, J., Gao, J., Li, Y.: Backward problem for a time-space fractional diffusion equation. Inverse Probl. Imaging 12, 773–799 (2018)
Jiang, D., Li, Z., Liu, Y., Yamamoto, M.: Weak unique continuation property and a related inverse source problem for time-fractional diffusion-advection equations. Inverse Probl. 33, 055013 (2017)
Jin, L., Li, L., Fang, S.: The global existence and time-decay for the solutions of the fractional pseudo-parabolic equation. Comput. Math. Appl. 73, 2221–2232 (2017)
Jothilakshmi, G., Vadivoo, B.S., Almalki, Y., Debbouche, A.: Controllability analysis of multiple fractional order integro-differential damping systems with impulsive interpretation. J. Comput. Appl. Math. 410, 114204 (2022)
Kaltenbacher, B., Rundell, W.: On an inverse potential problem for a fractional reaction–diffusion equation. Inverse Probl. 35, 065004 (2019)
Kaltenbacher, B., Rundell, W.: Regularization of a backward parabolic equation by fractional operators. Inverse Probl. Imaging 13, 401–430 (2019)
Kilbas, A.A., Saigo, M., Saxena, R.K.: Generalized Mittag–Leffler function and generalized fractional calculus operators. Integral Transforms Spec. Funct. 15(1), 31–49 (2004)
Kumar, V., Muslim, M., Amar, D.: Total controllability of neutral fractional differential equation with non-instantaneous impulsive effects. J. Comput. Appl. Math. 383, 113158 (2004)
Li, Z., Imanuvilov, O.Y., Yamamoto, M.: Uniqueness in inverse boundary value problems for fractional diffusion equations. Inverse Probl. 32, 015004 (2015)
Li, G., Zhang, D., Jia, X., Yamamoto, M.: Simultaneous inversion for the space-dependent diffusion coefficient and the fractional order in the time-fractional diffusion equation. Inverse Probl. 29, 065014 (2013)
Luchko, Y., Rundell, W., Yamamoto, M., Zuo, L.: Uniqueness and reconstruction of an unknown semilinear term in a time-fractional reaction–diffusion equation. Inverse Probl. 29, 065019 (2013)
Minh, N.D., Khanh, T.D., Tuan, N.H., Trong, D.D.: A two-dimensional backward heat problem with statistical discrete data. J. Inverse Ill-Posed Probl. 26, 13–31 (2018)
Nane, E., Tuan, N.H.: Approximate solutions of inverse problems for nonlinear space fractional diffusion equations with randomly perturbed data. SIAM J. Uncertain. Quantif. 6, 302–338 (2018)
Nguyen, H.T., Le, D.L., Nguyen, V.T.: Regularized solution of an inverse source problem for a time fractional diffusion equation. Appl. Math. Model. 40, 8244–8264 (2016)
Rundell, W., Zhang, Z.: Recovering an unknown source in a fractional diffusion problem. J. Comput. Phys. 368, 299–314 (2018)
Rundell, W., Zhang, Z.: Fractional diffusion: recovering the distributed fractional derivative from overposed data. Inverse Probl. 33, 27 (2017)
Subbaiyan, S., Amar, D., Jin, R.W.: Approximate controllability of Hilfer fractional Sobolev type integrodifferential inclusions with nonlocal conditions. Int. J. Dyn. Syst. Differ. Equ. 10, 59–80 (2020)
Tuan, N.H., Thach, T.N., Zhou, Y.: On a backward problem for two-dimensional time fractional wave equation with discrete random data. Evol. Equ. Control Theory 9, 561–579 (2020)
Tuan, N.H., Baleanu, D., Thach, T.N., O’Regan, D., Can, N.H.: Final value problem for nonlinear time fractional reaction–diffusion equation with discrete data. J. Comput. Appl. Math. 376, 112883 (2020)
Tuan, N.A., O’Regan, D., Baleanu, D., Tuan, N.H.: On time fractional pseudo-parabolic equations with nonlocal integeral conditions. Evol. Equ. Control Theory 11, 225 (2022)
Tuan, N.H., Long, L.D., Thinh, V.N., Thanh, T.: On a final value problem for the time-fractional diffusion equation with inhomogeneous source. Inverse Probl. Sci. Eng. 25, 1367–1395 (2017)
Tuan, N.H., Caraballo, T.: On initial and terminal value problems for fractional nonclassical diffusion equations. Proc. Am. Math. Soc. 149, 143–161 (2021)
Tuan, N.H., Amar, D., Tran, B.N.: Existence and regularity of final value problems for time fractional wave equations. Comput. Math. Appl. 78, 1396–1414 (2019)
Triet, N.A., Binh, T.T., Phuong, N.D., Baleanu, D., Can, N.H.: Recovering the initial value for a system of nonlocal diffusion equations with random noise on the measurements. Math. Methods Appl. Sci. 44, 5188–5209 (2021)
Triet, N.A., Tuan, N.H., Phuong, N.D., O’Regan, D.: On the inverse problem for nonlinear strongly damped wave equations with discrete random noise. Int. J. Nonlinear Sci. Numer. Simul. 23, 365–383 (2022)
Tuan, N.H., Nguyen, A.T., Yang, C.: Global well-posedness for fractional Sobolev–Galpern type equations. Discrete Contin. Dyn. Syst. 42, 2637–2665 (2022)
Tuan, N.H., Au, V.V., Nguyen, A.T.: Mild solutions to a time-fractional Cauchy problem with nonlocal nonlinearity in Besov spaces. Arch. Math. (Basel) 118, 305–314 (2022)
Tuan, N.H., Foondun, M., Thach, T.N., Wang, R.: On backward problems for stochastic fractional reaction equations with standard and fractional Brownian motion. Bull. Sci. Math. 179, 58 (2022)
Wang, L., Liu, J.: Total variation regularization for a backward time-fractional diffusion problem. Inverse Probl. 29, 115013 (2013)
Xu, J., Zhang, Z., Caraballo, T.: Mild solutions to time fractional stochastic 2D-Stokes equations with bounded and unbounded delay. J. Dyn. Differ. Equ. 34, 583–603 (2022)
Xu, J., Zhang, Z., Caraballo, T.: Non-autonomous nonlocal partial differential equations with delay and memory. J. Differ. Equ. 270, 505–546 (2021)
Zhao, C., Caraballo, T., Lukaszewicz, G.: Statistical solution and Liouville type theorem for the Klein–Gordon–Schrödinger equations. J. Differ. Equ. 281, 1–32 (2021)
Zhao, C., Wang, J., Caraballo, T.: Invariant sample measures and random Liouville type theorem for the two-dimensional stochastic Navier–Stokes equations. J. Differ. Equ. 317, 474–494 (2022)
Acknowledgements
The third author wishes to thank Faculty of Fundamental Science of Industrial University of Ho Chi Minh City, Vietnam, for the opportunity to work in it.
Funding
Open access funding provided by Università degli Studi di Catania within the CRUI-CARE Agreement.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Rosihan M. Ali.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Le Dinh, L., Nguyen, D.P. & Ragusa, M.A. On a Non-local Sobolev–Galpern-Type Equation Associated with Random Noise. Bull. Malays. Math. Sci. Soc. 46, 202 (2023). https://doi.org/10.1007/s40840-023-01595-y
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s40840-023-01595-y