
The Kuramoto-Sinelshchikov equation describes the evolution of a phase turbulence in reaction-diffusion systems or the evolution of the plane flame propagation, taking in account the combined influence of diffusion and thermal conduction of the gas on the stability of a plane flame front. In this paper, we prove the well-posedness of the classical solutions for the initial-boundary value problem for this equation, under appropriate boundary conditions.
Citation: Giuseppe Maria Coclite, Lorenzo di Ruvo. On the initial-boundary value problem for a Kuramoto-Sinelshchikov type equation[J]. Mathematics in Engineering, 2021, 3(4): 1-43. doi: 10.3934/mine.2021036
[1] | Yu Xu, Youjun Deng, Dong Wei . Numerical solution of forward and inverse problems of heat conduction in multi-layered media. AIMS Mathematics, 2025, 10(3): 6144-6167. doi: 10.3934/math.2025280 |
[2] | Fan Yang, Qianchao Wang, Xiaoxiao Li . A fractional Landweber iterative regularization method for stable analytic continuation. AIMS Mathematics, 2021, 6(1): 404-419. doi: 10.3934/math.2021025 |
[3] | Dun-Gang Li, Fan Yang, Ping Fan, Xiao-Xiao Li, Can-Yun Huang . Landweber iterative regularization method for reconstructing the unknown source of the modified Helmholtz equation. AIMS Mathematics, 2021, 6(9): 10327-10342. doi: 10.3934/math.2021598 |
[4] | Zhongdi Cen, Jian Huang, Aimin Xu, Anbo Le . An iterative method for solving a PDE with free boundary arising from pricing corporate bond with credit rating migration. AIMS Mathematics, 2023, 8(2): 3286-3302. doi: 10.3934/math.2023169 |
[5] | Imo Kalu Agwu, Umar Ishtiaq, Naeem Saleem, Donatus Ikechi Igbokwe, Fahd Jarad . Equivalence of novel IH-implicit fixed point algorithms for a general class of contractive maps. AIMS Mathematics, 2023, 8(1): 841-872. doi: 10.3934/math.2023041 |
[6] | Xiaopeng Yi, Chongyang Liu, Huey Tyng Cheong, Kok Lay Teo, Song Wang . A third-order numerical method for solving fractional ordinary differential equations. AIMS Mathematics, 2024, 9(8): 21125-21143. doi: 10.3934/math.20241026 |
[7] | Christelle Dleuna Nyoumbi, Antoine Tambue . A fitted finite volume method for stochastic optimal control problems in finance. AIMS Mathematics, 2021, 6(4): 3053-3079. doi: 10.3934/math.2021186 |
[8] | Xu Li, Rui-Feng Li . Shift-splitting iteration methods for a class of large sparse linear matrix equations. AIMS Mathematics, 2021, 6(4): 4105-4118. doi: 10.3934/math.2021243 |
[9] | Shubham Kumar Mittal, Sunil Panday, Lorentz Jäntschi, Liviu C. Bolunduţ . Two novel efficient memory-based multi-point iterative methods for solving nonlinear equations. AIMS Mathematics, 2025, 10(3): 5421-5443. doi: 10.3934/math.2025250 |
[10] | Mouhamad Al Sayed Ali, Miloud Sadkane . Acceleration of implicit schemes for large systems of nonlinear differential-algebraic equations. AIMS Mathematics, 2020, 5(1): 603-618. doi: 10.3934/math.2020040 |
The Kuramoto-Sinelshchikov equation describes the evolution of a phase turbulence in reaction-diffusion systems or the evolution of the plane flame propagation, taking in account the combined influence of diffusion and thermal conduction of the gas on the stability of a plane flame front. In this paper, we prove the well-posedness of the classical solutions for the initial-boundary value problem for this equation, under appropriate boundary conditions.
In the application of many mathematical physics problems, we need to estimate derivatives of an unknown function from given noisy data. It turns out to be an ill-posed problem, which means, the small errors in the measurement data can induce huge errors in its computed derivatives. Many methods and techniques have been proposed regarding this topic [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17]. According to type of regularization techniques, these methods can be classified into finite difference methods, mollification methods, differentiation by integration method and Tikhonov methods. Most of these methods are for one-dimensional case, but only a few are for high-dimensional cases [18,19,20]. These methods of dealing with two-dimensional problems are basically aimed at the regular region, and there are few methods that can deal with the numerical differentiation problem on the irregular domain. For example, if the data are given at scatter points of the domain in Figure 1, numerical implementation of most existing methods is difficult.
By the extension theorem [21], any function on sub-domain of R2 can be extended to R2 without losing its smoothness. There are many work of extension problem has been developed since Whitney's seminal work [22,23,24,25]. And it is well known that functions with high smoothness can be approximated very precisely by their Hermite expansion. So in this paper, we consider a Hermite extension method to deal with the numerical differentiation problem on the irregular domain. That is to say, we take a function defined on any domain as part of a function on R2. For the one-dimensional case, the numerical differentiation method along the line of this method has been given in [26]. The numerical process of extension is usually unstable, so regularization technology is needed. As an alternative regularization method, we use a modified implicit iteration method in this paper. Compared with Tikhonov method which is used in [17], the implicit iteration method can select larger regularization parameters in each iteration, which makes the calculation process more stable. In [27], Jin has given the detailed theory and numerical implementation of implicit iteration method in Hilbert scales. But the application of the method to our problem still needs some improvement. The existing implicit iterative method can only deal with the problem of finite smoothness, and the numerical implementation of the algorithm is inconsistent for different smoothness. In this paper, we will present a modified form of implicit iteration method. It can deal with numerical differentiation of functions with any smoothness and the solution process is uniform.
This paper is organized as follows: In the next section, we introduce some preliminary materials. In Section 3, we describe the modified implicit iteration method for numerical differentiation problem and give some auxiliary results. The convergence estimate of the approximation solution can be founded in Section 4. In Section 5, some numerical examples are given. Conclusion is then given in the final Section 6.
In this section, we introduce some notations and preliminaries that will be used throughout the paper.
Let x=(x1,x2), |x|=√x21+x22. Let
ˆf(ω)=F[f(x)]=12π∫R2f(x)e−iω⋅xdx |
be the 2-dimensional Fourier transform of the function f(x)∈L2(R2). The corresponding inverse Fourier transform of the function ˆf(ω) is
f(x)=12π∫R2ˆf(ω)eiω⋅xdω. |
And ‖⋅‖p denotes the norm of the Sobolev space Hp(R2) with p≥0 defined by
‖f‖p=(∫R2(1+|ω|2)p|ˆf(ω)|2dω)12. | (2.1) |
Particularly, for p=0 we can recover the L2(R2) norm, i.e.,
‖f‖=(∫R2|ˆf(ω)|2dω)12. |
Let ℓ=(ℓ1,ℓ2), ℓi(i=1,2) being non-negative integers. Set |ℓ|1=ℓ1+ℓ2 and ℓ⋅x=ℓ1x1+ℓ2x2. The normalized 2-dimensional Hermite function is defined by
Hℓ(x)=Hℓ1(x1)Hℓ2(x2), | (2.2) |
where
H0(xi)≡π−1/4exp(−(1/2)x2i),H1(xi)≡π−1/4√2xiexp(−(1/2)x2i),Hℓi+1(xi)=√2ℓi+1xiHℓi(xi)−√ℓiℓi+1Hℓi−1(xi),ℓi≥1. | (2.3) |
We know that the Fourier transform of the Hermite functions can be given as[28]:
ˆHℓ(ω)=(−i)|ℓ|1Hℓ(ω). | (2.4) |
The set of Hermite functions satisfy the orthogonality relations
∫R2Hℓ(x)Hm(x)dx=δℓ,m. | (2.5) |
The Hermite expansion of a function f∈L2(R2) is as
f(x)=∞∑|ℓ|1=0fℓHℓ(x), | (2.6) |
with the Fourier-Hermite coefficients
fℓ=∫R2f(x)Hℓ(x)dx,|ℓ|1=0,1,…. | (2.7) |
Suppose that Λ is a subdomain of R2. For any two-tuples α=(α1,α2)∈N2,|α|1=α1+α2. The notation ∂j stands for ∂∂xj and Dαf=∂α1x1∂α2x2f. For any positive integer q, we define Dqf=:{Dαf:|α|1=q} and
|Dqf|=(∑|α|1=q|Dαf|2)1/2. | (2.8) |
The norm ‖⋅‖s,Λ in Sobolev space Hs(Λ) is defined as
‖f‖s,Λ:=(∫Λ|f|2+|Dsf|2dx)1/2, | (2.9) |
where s=0 and ‖⋅‖0,Λ denotes the L2(Λ) norm.
Suppose that g(x)∈Hp(R),p≥2 and we only know its approximate function gδ on Λ such that
‖gδ−g‖0,Λ≤δ, | (2.10) |
where δ>0 is a given constant called the error level. Our problem is to calculate approximate derivatives of g on Λ from the noisy data gδ, or, equivalently, to construct a function fδ(x) from gδ(x) which is close to g(x) in the sense that
limδ→0‖fδ−g‖r,Λ=0,r≥1. | (2.11) |
For any vector →f={fℓ}∞|ℓ|1=0∈l2, if we let
H→f:=∞∑|ℓ|=0fℓHℓ(x), | (3.1) |
then the process of constructing an approximation function fδ from data gδ can be transformed to solving the following equations
H→f=gδ. | (3.2) |
In this paper, we present an modified implicit iteration method to solve the above equations. For this purpose, we introduce the following operator:
R→f:=H−1F−1[e|ω|^H→f(ω)]. | (3.3) |
It is obvious that R is unbounded self-adjoint strictly positive definite operator. Then we choose
fδn=H→fδn, | (3.4) |
as the approximation of g, where →fδn is determined by the following implicit iteration process
→fδ0=0,→fδk=→fδk−1−(H∗H+βkR2)−1H∗(H→fδk−1−gδ),k=1,2,…,n, | (3.5) |
where βk>0 are properly chosen real numbers. For reference [27], the positive number
σn:=n∑k=11βk | (3.6) |
plays the role of the regularization parameter and we will chosen it as the solution of the nonlinear equation
d(σn):=‖H→fδn−gδ‖0,Λ=Cδ, | (3.7) |
with a constant C≥1. If we let T=HR−1, then →fδn possesses the representation [27]
→fδn=R−1sn(T∗T)T∗gδwithsn(λ)=1λ(1−n∏k=1βkλ+βk). | (3.8) |
Remark 3.1. If we use the operator B=(∑|α|1=qDα) with some a constant q instead of R, then we return to the framework in [27] and the convergence results can be obtained accordingly. When p and q satisfy a certain relation, the result is order optimal. It should be noticed that for large q, the numerical process of the method is difficult. We will point that the method is always order optimal when we use the operator R and the numerical process is uniform for any p.
The following lemma holds for sn(λ).
Lemma 3.1. [27] The function sn:(0,c]→(0,∞) with c=‖T‖2 and the corresponding residual function rn(λ):=1−λsn(λ) obey the properties
sn(λ)≤σn,λsn(λ)≤1,λrn(λ)≤σ−1n,rn(λ)≤1. | (3.9) |
From above lemma, we can deduce the following results.
Lemma 3.2.
√λsn(λ)≤√σn,√λrn(λ)≤√σ−1n. | (3.10) |
Proof. For λ≤σ−1,
√λsn(λ)≤√λσn≤√σn | (3.11) |
and
√λrn(λ)≤√λ≤√σ−1n. | (3.12) |
Moreover, for λ≥σ−1,
√λsn(λ)=√λλλsn(λ)≤√λλ≤√σn | (3.13) |
and
√λrn(λ)≤√λλλrn(λ)≤σ−1n√λ≤√σ−1n. | (3.14) |
Owing to g∈Hp(R), we suppose that
‖g‖p≤E, | (4.1) |
where E is a constant. Set the vector →g contains all Fourier-Hermite coefficients of g, i.e.,
g(x)=(H→g)(x),∀x∈R2. | (4.2) |
Let
→gN=PN→gandgN=H→gN. | (4.3) |
We define the vector →fn as
→fn=R−1sn(T∗T)T∗gN, | (4.4) |
then we have
H(→fδn−→fn)=Tsn(T∗T)T∗(gδ−gN), | (4.5) |
H(→g−→fn)=Trn(T∗T)R→gN, | (4.6) |
gδ−H→fδn=rn(TT∗)gδ, | (4.7) |
R(→fδn−→fn)=sn(T∗T)T∗(gδ−gN), | (4.8) |
R(→g−→fn)=rn(T∗T)R→gN. | (4.9) |
In our further analysis, we shall make use of the following lemmas.
Lemma 4.1. If the condition (4.1) holds, then
‖g−gN‖≤N−pEand‖R→gN‖l2≤CNE, | (4.10) |
where
CN=max(1,eNNp). | (4.11) |
Proof. From (2.1) and (3.3), we can obtain
‖g−gN‖2=∫|ω|>N|ˆg(ω)|2dω≤N−2p∫|ω|>N(1+|ω|2)p|ˆg|2dω≤N−2p‖g‖2p | (4.12) |
and
‖R→gN‖l2=∫|ω|≤Ne2|ω||ˆg(ω)|2dω=∫|ω|≤Ne2|ω|(1+|ω|2)p(1+|ω|2)p|ˆg(ω)|2dω≤max(1,e2NN2p)‖g‖2p. | (4.13) |
Lemma 4.2. If the condition (4.1) holds, we have
‖H(→fδn−→gN)‖0,Λ≤(C+1)δ+N−pE, | (4.14) |
‖R(→fδn−→gN)‖l2≤√σn(δ+N−pE)+CNE | (4.15) |
and
‖H→fδn−gδ‖0,Λ≤δ+N−pE+√σ−1nCNE. | (4.16) |
Proof. From (2.10), (3.7), (4.10) and the triangle inequality
‖H(→fδn−→gN)‖0,Λ≤‖H→fδn−gδ‖0,Λ+‖gδ−g‖0,Λ+‖g−gN‖0,Λ≤(C+1)δ+N−pE. | (4.17) |
And by using the triangle inequality, (2.10), (3.10) and (4.8)–(4.10)
‖R(→fδn−→gN)‖l2≤‖R(→fδn−→fn)‖l2+‖R(→fn−→gN)‖l2=‖sn(T∗T)T∗(gδ−gN)‖l2+‖rn(T∗T)R→gN‖l2≤√σn‖(gδ−gN)‖0,Λ+‖R→gN‖l2≤√σn(δ+N−pE)+CNE. | (4.18) |
Moreover, in terms of the triangle inequality, (2.10), (3.9), (3.10) and (4.7), we have
‖H→fδn−gδ‖0,Λ=‖rn(TT∗)gδ‖0,Λ≤‖rn(TT∗)(gδ−g)‖0,Λ+‖rn(TT∗)(g−gN)‖0,Λ+‖rn(TT∗)gN‖0,Λ≤δ+‖g−gN‖0,Λ+‖rn(TT∗)T‖⋅‖R→gN‖≤δ+N−pE+√σ−1nCNE. | (4.19) |
Lemma 4.3. [21] Let Ω be a domain in R2 satisfying the cone condition. There exists a constant K depending on ϵ0 and j, s, such that for any 0<ϵ≤ϵ0 and 0≤j≤s
‖f‖j,Ω≤K(ϵ‖f‖s,Ω+ϵ−j/(s−j)‖f‖0,Ω). | (4.20) |
Lemma 4.4. Suppose that the vector sequence →hδn={hδℓ}∞|ℓ|1=0 satisfies
‖H→hδ‖0,Λ≤k1δ,‖R→hδ‖l2≤k2ek3δ−1pδ,δ→0, | (4.21) |
then for any Ω⊆Λ satisfying the cone condition, there exists a constant M
‖H→hδ‖p,Ω≤M. | (4.22) |
Proof. It is easy to deduce that there exist a constant δ0 such that
ek3δ−1p>kp3δ,∀δ<δ0. | (4.23) |
And for simplicity, we prove the theorem with δ<δ0. Let
N0=k3δ−1p, | (4.24) |
and we have
‖H→hδ‖p,Ω≤‖H(PN0→hδ)‖p,Ω+‖H[(I−PN0)→hδ]‖p,Ω=I1+I2. | (4.25) |
By Parseval's formula, we can see that the second term I2 satisfies
‖H[(I−PN0)→hδ]‖2p,Ω≤‖H[(I−PN0)→hδ]‖2=∫|ω|>N0(1+|ω|2)p|^H→hδ(ω)|2dω=∫|ω|>N0(1+|ω|2)pe2|ω||e2|ω|^H→hδ(ω)|2dω≤(N0+1)2pe2N0∫|ω|>N0|e2|ω|^H→hδ(ω)|2dω≤N2p0e2(N0−1)‖R→hδ‖2l2≤e−2k2p31δ2⋅k22δ2=e−2k2p3k22. | (4.26) |
Hence
I2≤2e−1kp3k2. | (4.27) |
So all we need is to prove there exist a constant M1 such that
I1<M1,δ→0. | (4.28) |
Note that
‖H(PN0→hδ)‖0,Ω≤‖H→hδ‖0,Ω+‖H[(I−PN0)→hδ]‖0,Ω | (4.29) |
and
‖H(I−PN0)→hδ‖0,Ω≤‖H[(I−PN0)→hδ]‖2=∫|ω|>N0|^H→hδ(ω)|2dω=∫|ω|>N01e2|ω||e|ω|^H→hδ(ω)|2dω≤1e2N0‖R→hδ‖2l2≤k22δ2. | (4.30) |
Therefore
‖H(PN0→hδ)‖0,Ω≤(k1+2k2)δ. | (4.31) |
Now we prove (4.28) by using reduction to absurdity, if (4.28) does not hold, then for any q>p there exist a sequence δi such that
‖H(PN0→hδ)‖q≥2k2kp3(δikp3)p−qp,δi→0. | (4.32) |
If not, ∃ˉq for any δ→0
‖H(PN0→hδ)‖ˉq<2k2kp3(δkp3)p−ˉqp, | (4.33) |
then (4.28) can be derived by 4.29 and Lemma 4.3 with ϵ=(δikp3)ˉq−pp, s=ˉq and j=p. Then
∫|ω|<N0(N0∑k=0|ω|kk!)2|^H→hδi(ω)|2dω=∫|ω|<N0(N0∑k=0|ω|kk!)2(1+|ω|2)N0(1+|ω|2)N0|^H→hδi(ω)|2dω≥∫|ω|<N0(N0∑k=0|ω|kk!)2(1+|ω|)2N0(1+|ω|2)N0|^H→hδi(ω)|2dω≥∫|ω|<N0(N0−1∑k=0|ω|kk!)2|ω|2N0(1+|ω|2)N0|^H→hδi(ω)|2dω≥∫|ω|<N0(N0−1∑k=0N0kk!)2N2N00(1+|ω|2)N0|^H→hδi(ω)|2dω≥(N0−1∑k=0N0kk!)2N2N00‖H(PN0→hδi)‖2N0≥k22(N0∑k=0(k3δ−1pi)kk!)2δ2i. | (4.34) |
Therefore
∫R2e2|ω||^H→hδi(ω)|2dω=limδi→0∫|ω|<N0(δi)(N0(δi)∑k=0|ω|kk!)2|^H→hδi(ω)|2dω≥k22(N0∑k=0(k3δ−1pi)kk!)2δ2i=k22limδi→0e2k3δ−1piδ2i. | (4.35) |
So there exists a ˉδ such that
‖R→hˉδ‖2l2=∫R2e2|ω||^Hhˉδ(ω)|2dω>∫R2e2|ω||^Hhˉδ(ω)|2dω≥k22e2k3ˆδ−1pˉδ2, | (4.36) |
which contradicts the assumptions of the Lemma.
Theorem 4.1. Suppose that the conditions (2.10) and (4.1) hold, fδn is defined by (3.4) and (3.7) then for any Ω⊆Λ satisfying the cone condition and 0<j≤p,
‖fδn−g‖j,Ω=O(δp−jp). | (4.37) |
Proof. Let
N0=(2E(C−1)δ)1p. | (4.38) |
Then from Lemma 4.2, we have
‖H(→fδn−→gN0)‖0,Λ≤3C+12δ, | (4.39) |
‖R(→fδn−→gN0)‖l2≤C+12e(2EC−1)1pδ−1pδ. | (4.40) |
Thus, by using Lemma 4.4, there exists a constant M
‖H(→fδn−→gN0)‖0,Ω≤M. | (4.41) |
Then
‖H→fδn−g‖p,Ω≤‖H(→fδn−→gN0)‖0,Ω+‖g−gN‖p,Ω≤‖H(→fδn−→gN0)‖0,Ω+‖g‖p≤M+E. | (4.42) |
Moreover, by using (2.10), (3.7) and the triangle inequality
‖H→fδn−g‖0,Ω≤‖H→fδn−gδ‖0,Ω+‖gδ−g‖0,Ω≤‖H→fδn−gδ‖0,Λ+‖gδ−g‖0,Λ≤(C+1)δ. | (4.43) |
The assertion of theorem follows from (4.42), (4.43) and Lemma 4.3.
The data are usually given at scatter points in practical applications. Let xi∈Λ(i=1,2,…,m) be the given points and
gδ=(gδ(x1),gδ(x2),…,gδ(xm))T |
be the noisy data vector. Let σ(j)=(σ(j1),σ(j2)),(0≤j1,j2≤n) being the Hermite-Gauss type interpolation points and ρ(j) are the corresponding Hermite-Gauss weights. For f,g∈L2(R2), we define the discrete inner product
⟨f,g⟩n:=n∑j1=1n∑j2=1ρ(j)f(σ(j))¯g(σ(j)), |
and
˘Hℓ(ω):=(−1)|ℓ|1e|ω|Hℓ(ω). |
Let
Hn=span{H(0,0)(x),H(1,0)(x),H(1,1)(x),…,H(n,n−1)(x),H(n,n)(x)}, |
then we give the matrices A(n+1)2×(n+1)2, R(n+1)2×(n+1)2, Hm×(n+1)2 as
A|ℓ|1+1,|k|1+1=∑mi=1Hℓ(xi)Hk(xi),R|ℓ|1+1,|k|1+1=⟨˘Hℓ,˘Hk⟩n,Hi,|ℓ|1+1=Hℓ(xi),i=1,2,…,m;|ℓ|1,|k|1=0,1,…n. |
With these preparations, the discrete form of the implicit iteration method can be given as
fδ0=0,fk=fk−1−(HTH+βkR)−1HT(Hfδk−1−gδ),k=1,2,…,n. | (5.1) |
Suppose that
(m∑i=1(gδ(xi)−g(xi)2)1/2≤δ. | (5.2) |
Similar to what is done in [27], we take β1=1,βk=qk−1β1 with some q<1 and choose n as the first integer for which
‖Hfδn−gδ‖≤C1δ<‖Hfδk−gδ‖,0≤k<n, | (5.3) |
and then adjust the parameter βn such that
C2δ≤‖Hfδn−gδ‖≤C1δ, | (5.4) |
where C1,C2 are two constants that obey 1≤C1≤C2.
Remark 5.1. It should be noted that the Hermite-Gauss points are only used to calculate the matrix R, regardless of the location of the noisy data.
In this section, we give some numerical tests to verify the effect of the new method. All tests are realized on Windows 10 system with Memory 16GB, CPU Intel(R) Core(TM)i7-8550U by using Matlab 2017b. Let x=(x1,x2,…,xm)T and the perturbed data are generated by
gδ(x)=g(x)+randn(size(x))⋅ϵ, | (6.1) |
where ϵ is the error level and randn(size(⋅)) is Matlab functions. In all cases we choose the parameter n=64, q=1/2 and C=1.01. We have tested these parameters with other values, and the results are similar. In order to adapt to the characteristic of Hermite function approximation, the scaling factor [29] is used in the numerical processing.
Example 6.1. [18] Let Λ={(x1,x2)|x21+x22≤1} is a disk and scatter nodes are given as Figure 2a. We choose the exact function as g(x)=(x21+x22−2)3 and set ϵ=0.01. The numerical results are exhibited in Figures 2c–3h.
Example 6.2. [18] Let Λ={(x1,x2)|0≤x1≤1,0≤x2≤π} is a rectangle and scatter nodes are given as Figure 3i. We choose the exact function as g(x)=(x21+x22−2)3 and set ϵ=0.01. We have shown the numerical results in Figures 3k–4h.
Example 6.3. [19] Let Λ={(x1,x2)|−1≤x1≤3,−1≤x2≤3}, the data are given at the equidistant nodes whose sampling step was 0.1×0.1. The exact function is chosen as g(x)=sin(12x21+14x22+3)cos(2x1+1−exp(x2)). We have shown the numerical results in Figures 4i–5h with ϵ=0.01.
Example 6.4. Now we let Λ is a irregular domain and scatter nodes are given as Figure 5i. We choose the exact function as g(x)=cos(x1⋅x2) and set ϵ=0.01. The numerical approximations and corresponding errors are shown in Figures 5k–6h.
All the above numerical results show that the proposed method is effective.
In this paper, we present a Hermite extension method with an implicit iteration process for numerical differentiation of two-dimensional functions. Because the method can directly deal with the data given on any domain, it is more convenient than other methods in practical application. The theoretical results show that the convergence rates of the method is self-adaptive.
The authors declared that they have no conflicts of interest to this work. We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.
[1] |
Amiranashvili S, Vladimirov AG, Bandelow U (2008) Solitary-wave solutions for few-cycle optical pulses. Phys Rev A 77: 063821. doi: 10.1103/PhysRevA.77.063821
![]() |
[2] |
Amiranashvili S, Vladimirov AG, Bandelow U (2010) A model equation for ultrashort optical pulses around the zero dispersion frequency. Eur Phys J D 58: 219-226. doi: 10.1140/epjd/e2010-00010-3
![]() |
[3] |
Armaou A, Christofides PD (2000) Feedback control of the Kuramoto-Sivashinsky equation. Phys D 137: 49-61. doi: 10.1016/S0167-2789(99)00175-X
![]() |
[4] |
Benney DJ (1966) Long waves on liquid films. J Math Phys 45: 150-155. doi: 10.1002/sapm1966451150
![]() |
[5] | Biagioni HA, Bona JL, Iório Jr RJ, et al. (1996) On the Korteweg-de Vries-Kuramoto-Sivashinsky equation. Adv Differ Equ 1: 1-20. |
[6] |
Cerpa E (2010) Null controllability and stabilization of the linear Kuramoto-Sivashinsky equation. Commun Pure Appl Anal 9: 91-102. doi: 10.3934/cpaa.2010.9.91
![]() |
[7] | Chen LH, Chang HC (1986) Nonlinear waves on liquid film surfaces-II. Bifurcation analyses of the long-wave equation. Chem Eng Sci 41: 2477-2486. |
[8] |
Christofides PD, Armaou A (2000) Global stabilization of the Kuramoto-Sivashinsky equation via distributed output feedback control. Syst Control Lett 39: 283-294. doi: 10.1016/S0167-6911(99)00108-5
![]() |
[9] |
Coclite GM, di Ruvo L (2014) Convergence of the Ostrovsky equation to the Ostrovsky-Hunter one. J Differ Equations 256: 3245-3277. doi: 10.1016/j.jde.2014.02.001
![]() |
[10] |
Coclite GM, di Ruvo L (2015) Dispersive and diffusive limits for Ostrovsky-Hunter type equations. NoDEA Nonlinear Diff 22: 1733-1763. doi: 10.1007/s00030-015-0342-1
![]() |
[11] |
Coclite GM, di Ruvo L (2015) Well-posedness of bounded solutions of the non-homogeneous initial-boundary value problem for the Ostrovsky-Hunter equation. J Hyperbolic Differ Equ 12: 221-248. doi: 10.1142/S021989161550006X
![]() |
[12] |
Coclite GM, di Ruvo L (2016) Convergence of the Kuramoto-Sinelshchikov equation to the Burgers one. Acta Appl Math 145: 89-113. doi: 10.1007/s10440-016-0049-2
![]() |
[13] |
Coclite GM, di Ruvo L (2016) Convergence of the solutions on the generalized Korteweg-de Vries equation. Math Model Anal 21: 239-259. doi: 10.3846/13926292.2016.1150358
![]() |
[14] | Coclite GM, di Ruvo L (2017) A singular limit problem for conservation laws related to the Rosenau equation. J Abstr Differ Equ Appl 8: 24-47. |
[15] |
Coclite GM, di Ruvo L (2019) Discontinuous solutions for the generalized short pulse equation. Evol Equ Control The 8: 737-753. doi: 10.3934/eect.2019036
![]() |
[16] | Coclite GM, di Ruvo L (2020) Convergence of the Rosenau-Korteweg-de Vries equation to the Korteweg-de Vries one. Contemporary Mathematics. |
[17] | Coclite GM, di Ruvo L (2020) A note on the non-homogeneous initial boundary problem for an Ostrovsky-Hunter1 type equation. Discrete Contin Dyn Syst Ser S 13: 3357-3389. |
[18] |
Coclite GM, di Ruvo L (2020) On classical solutions for a Kuramoto-Sinelshchikov-Velarde-type equation. Algorithms 13: 77. doi: 10.3390/a13040077
![]() |
[19] |
Coclite GM, di Ruvo L (2020) On the solutions for an Ostrovsky type equation. Nonlinear Anal Real 55: 103141. doi: 10.1016/j.nonrwa.2020.103141
![]() |
[20] |
Cohen BI, Krommes JA, Tang WM, et al. (1976) Non-linear saturation of the dissipative trappedion mode by mode coupling. Nucl Fusion 16: 971-992. doi: 10.1088/0029-5515/16/6/009
![]() |
[21] | Foias C, Nicolaenko B, Sell GR, et al. (1988) Inertial manifolds for the Kuramoto-Sivashinsky equation and an estimate of their lowest dimension. J Math Pure Appl 67: 197-226. |
[22] |
Giacomelli L, Otto F (2005) New bounds for the Kuramoto-Sivashinsky equation. Commun Pure Appl Math 58: 297-318. doi: 10.1002/cpa.20031
![]() |
[23] |
Hooper AP, Grimshaw R (1985) Nonlinear instability at the interface between two viscous fluids. Phys Fluids 28: 37-45. doi: 10.1063/1.865160
![]() |
[24] | Hu C, Temam R (2001) Robust control of the Kuramoto-Sivashinsky equation. Dyn Contin Discrete Impuls Syst Ser B Appl Algorithms 8: 315-338. |
[25] |
Kenig CE, Ponce G, Vega L (1993) Well-posedness and scattering results for the generalized Korteweg-de Vries equation via the contraction principle. Commun Pure Appl Math 46: 527-620. doi: 10.1002/cpa.3160460405
![]() |
[26] | Khalique C (2012) Exact solutions of the generalized kuramoto-sivashinsky equation. CJMS 1: 109-116. |
[27] | Korteweg DDJ, de Vries DG (1895) XLI. On the change of form of long waves advancing in a rectangular canal, and on a new type of long stationary waves. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 39: 422-443. |
[28] |
Kudryashov NA (1990) Exact solutions of the generalized Kuramoto-Sivashinsky equation. Phys Lett A 147: 287-291. doi: 10.1016/0375-9601(90)90449-X
![]() |
[29] |
Kudryashov NA (2009) On "new travelling wave solutions" of the KdV and the KdV-Burgers equations. Commun Nonlinear Sci Numer Simul 14: 1891-1900. doi: 10.1016/j.cnsns.2008.09.020
![]() |
[30] |
Kuramoto Y (1978) Diffusion-induced chaos in reaction systems. Prog Theor Phys Supp 64: 346- 367. doi: 10.1143/PTPS.64.346
![]() |
[31] |
Kuramoto Y, Tsuzuki T (1975) On the formation of dissipative structures in reaction-diffusion systems: Reductive perturbation approach. Prog Theor Phys 54: 687-699. doi: 10.1143/PTP.54.687
![]() |
[32] |
Kuramoto Y, Tsuzuki T (1976) Persistent propagation of concentration waves in dissipative media far from thermal equilibrium. Prog Theor Phys 55: 356-369. doi: 10.1143/PTP.55.356
![]() |
[33] |
LaQuey RE, Mahajan SM, Rutherford PH, et al. (1975) Nonlinear saturation of the trapped-ion mode. Phys Rev Lett 34: 391-394. doi: 10.1103/PhysRevLett.34.391
![]() |
[34] |
Leblond H, Mihalache D (2009) Few-optical-cycle solitons: Modified Korteweg-de vries sinegordon equation versus other non-slowly-varying-envelope-approximation models. Phys Rev A 79: 063835. doi: 10.1103/PhysRevA.79.063835
![]() |
[35] |
Leblond H, Mihalache D (2013) Models of few optical cycle solitons beyond the slowly varying envelope approximation. Phys Rep 523: 61-126. doi: 10.1016/j.physrep.2012.10.006
![]() |
[36] |
Leblond H, Sanchez F (2003) Models for optical solitons in the two-cycle regime. Phys Rev A 67: 013804. doi: 10.1103/PhysRevA.67.013804
![]() |
[37] |
LeFloch PG, Natalini R (1999) Conservation laws with vanishing nonlinear diffusion and dispersion. Nonlinear Anal 36: 213-230. doi: 10.1016/S0362-546X(98)00012-1
![]() |
[38] | Li C, Chen G, Zhao S (2004) Exact travelling wave solutions to the generalized kuramotosivashinsky equation. Lat Am Appl Res 34: 65-68. |
[39] |
Li J, Zhang BY, Zhang Z (2017) A nonhomogeneous boundary value problem for the KuramotoSivashinsky equation in a quarter plane. Math Method Appl Sci 40: 5619-5641. doi: 10.1002/mma.4413
![]() |
[40] |
Li J, Zhang BY, Zhang Z (2020) A non-homogeneous boundary value problem for the KuramotoSivashinsky equation posed in a finite interval. ESAIM Control Optim Calc Var 26: 43. doi: 10.1051/cocv/2019027
![]() |
[41] |
Lin SP (1974) Finite amplitude side-band stability of a viscous film. J Fluid Mech 63: 417-429. doi: 10.1017/S0022112074001704
![]() |
[42] |
Liu WJ, Krstić M (2001) Stability enhancement by boundary control in the Kuramoto-Sivashinsky equation. Nonlinear Anal 43: 485-507. doi: 10.1016/S0362-546X(99)00215-1
![]() |
[43] |
Nicolaenko B, Scheurer B, Temam R (1985) Some global dynamical properties of the kuramotosivashinsky equations: Nonlinear stability and attractors. Physica D 16: 155-183. doi: 10.1016/0167-2789(85)90056-9
![]() |
[44] |
Nicolaenko B, Scheurer B (1984) Remarks on the kuramoto-sivashinsky equation. Physica D 12: 391-395. doi: 10.1016/0167-2789(84)90543-8
![]() |
[45] | Sajjadian M (2014) The shock profile wave propagation of Kuramoto-Sivashinsky equation and solitonic solutions of generalized Kuramoto-Sivashinsky equation. Acta Univ Apulensis Math Inform 38: 163-176. |
[46] |
Schonbek ME (1982) Convergence of solutions to nonlinear dispersive equations. Commun Part Diff Eq 7: 959-1000. doi: 10.1080/03605308208820242
![]() |
[47] | Sivashinsky G (1977) Nonlinear analysis of hydrodynamic instability in laminar flamesâ-I. Derivation of basic equations. Acta Astronaut 4: 1177-1206. |
[48] |
Tadmor E (1986) The well-posedness of the Kuramoto-Sivashinsky equation. SIAM J Math Anal 17: 884-893. doi: 10.1137/0517063
![]() |
[49] | Taylor ME (2011) Partial Differential Equations I. Basic Theory, 2 Eds., New York: Springer. |
[50] |
Topper J, Kawahara T (1978) Approximate equations for long nonlinear waves on a viscous fluid. J Phys Soc JPN 44: 663-666. doi: 10.1143/JPSJ.44.663
![]() |
[51] | Xie Y (2013) Solving the generalized Benney equation by a combination method. Int J Nonlinear Sci 15: 350-354. |
1. | Bin Zheng, Junfeng Liu, Zhenyu Zhao, Zhihong Dou, Benxue Gong, A Generalized Iterated Tikhonov Method in the Fourier Domain for Determining the Unknown Source of the Time-Fractional Diffusion Equation, 2024, 16, 2073-8994, 864, 10.3390/sym16070864 |