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

Divisibility among determinants of power matrices associated with integer-valued arithmetic functions

  • Let a,b and n be positive integers and S={x1,...,xn} be a set of n distinct positive integers. The set S is called a divisor chain if there is a permutation σ of {1,...,n} such that xσ(1)|...|xσ(n). We say that the set S consists of two coprime divisor chains if we can partition S as S=S1S2, where S1 and S2 are divisor chains and each element of S1 is coprime to each element of S2. For any arithmetic function f, we define the function fa for any positive integer x by fa(x):=(f(x))a. The matrix (fa(S)) is the n×n matrix having fa evaluated at the the greatest common divisor of xi and xj as its (i,j)-entry and the matrix (fa[S]) is the n×n matrix having fa evaluated at the least common multiple of xi and xj as its (i,j)-entry. In this paper, when f is an integer-valued arithmetic function and S consists of two coprime divisor chains with 1S, we establish the divisibility theorems between the determinants of the power matrices (fa(S)) and (fb(S)), between the determinants of the power matrices (fa[S]) and (fb[S]) and between the determinants of the power matrices (fa(S)) and (fb[S]). Our results extend Hong's theorem obtained in 2003 and the theorem of Tan, Lin and Liu gotten in 2011.

    Citation: Long Chen, Shaofang Hong. Divisibility among determinants of power matrices associated with integer-valued arithmetic functions[J]. AIMS Mathematics, 2020, 5(3): 1946-1959. doi: 10.3934/math.2020130

    Related Papers:

    [1] Gonca Kizilaslan . The altered Hermite matrix: implications and ramifications. AIMS Mathematics, 2024, 9(9): 25360-25375. doi: 10.3934/math.20241238
    [2] Hasan Gökbaş . Some properties of the generalized max Frank matrices. AIMS Mathematics, 2024, 9(10): 26826-26835. doi: 10.3934/math.20241305
    [3] Qin Zhong, Ling Li . Notes on the generalized Perron complements involving inverse N0-matrices. AIMS Mathematics, 2024, 9(8): 22130-22145. doi: 10.3934/math.20241076
    [4] Salima Kouser, Shafiq Ur Rehman, Mabkhoot Alsaiari, Fayyaz Ahmad, Mohammed Jalalah, Farid A. Harraz, Muhammad Akram . A smoothing spline algorithm to interpolate and predict the eigenvalues of matrices extracted from the sequence of preconditioned banded symmetric Toeplitz matrices. AIMS Mathematics, 2024, 9(6): 15782-15795. doi: 10.3934/math.2024762
    [5] Sumaira Hafeez, Rashid Farooq . On generalized inverse sum indeg index and energy of graphs. AIMS Mathematics, 2020, 5(3): 2388-2411. doi: 10.3934/math.2020158
    [6] Yuwen He, Jun Li, Lingsheng Meng . Three effective preconditioners for double saddle point problem. AIMS Mathematics, 2021, 6(7): 6933-6947. doi: 10.3934/math.2021406
    [7] Chih-Hung Chang, Ya-Chu Yang, Ferhat Şah . Reversibility of linear cellular automata with intermediate boundary condition. AIMS Mathematics, 2024, 9(3): 7645-7661. doi: 10.3934/math.2024371
    [8] Yongge Tian . Miscellaneous reverse order laws and their equivalent facts for generalized inverses of a triple matrix product. AIMS Mathematics, 2021, 6(12): 13845-13886. doi: 10.3934/math.2021803
    [9] Ahmad Y. Al-Dweik, Ryad Ghanam, Gerard Thompson, M. T. Mustafa . Algorithms for simultaneous block triangularization and block diagonalization of sets of matrices. AIMS Mathematics, 2023, 8(8): 19757-19772. doi: 10.3934/math.20231007
    [10] Wanlin Jiang, Kezheng Zuo . Revisiting of the BT-inverse of matrices. AIMS Mathematics, 2021, 6(3): 2607-2622. doi: 10.3934/math.2021158
  • Let a,b and n be positive integers and S={x1,...,xn} be a set of n distinct positive integers. The set S is called a divisor chain if there is a permutation σ of {1,...,n} such that xσ(1)|...|xσ(n). We say that the set S consists of two coprime divisor chains if we can partition S as S=S1S2, where S1 and S2 are divisor chains and each element of S1 is coprime to each element of S2. For any arithmetic function f, we define the function fa for any positive integer x by fa(x):=(f(x))a. The matrix (fa(S)) is the n×n matrix having fa evaluated at the the greatest common divisor of xi and xj as its (i,j)-entry and the matrix (fa[S]) is the n×n matrix having fa evaluated at the least common multiple of xi and xj as its (i,j)-entry. In this paper, when f is an integer-valued arithmetic function and S consists of two coprime divisor chains with 1S, we establish the divisibility theorems between the determinants of the power matrices (fa(S)) and (fb(S)), between the determinants of the power matrices (fa[S]) and (fb[S]) and between the determinants of the power matrices (fa(S)) and (fb[S]). Our results extend Hong's theorem obtained in 2003 and the theorem of Tan, Lin and Liu gotten in 2011.


    Nonlinear wave equations have been widely used to describe natural phenomena of science, engineering, geology, economics, meteorology, chemistry, and physics. The phenomena of dispersion, dissipation, diffusion, reaction and convection play a major role in nonlinear wave equations. Therefore, it is very important to find exact solutions of nonlinear evolution equations. Recently, there are many useful techniques to obtain the exact traveling wave solutions, such as the inverse scattering transform method [1], modified simple equation method [2], the Bäcklund transformation [3], the Lie symmetry method [4], the Darboux transformation [5], the multiple exp-function method [6], Hirota's bilinear method [7], the sine-cosine method [8], the bifurcation theory of dynamic system [9]. A well-known nonlinear wave equations is equal width (EW) equation

    ut+uuxbuxxt=0 (1.1)

    with the boundary conditions u0 as x±. Since it describes many physical phenomena such as shallow water waves and ion acoustic plasma waves, it is a model for the simulation of one-dimensional wave propagation in nonlinear media with dispersion processes [10]. The regularized long wave (RLW) equation is given by

    ut+ux+12(u2x)uxxt=0 (1.2)

    with <x<+, t>0. The RLW equation appears in many physical applications, for example, the nonlinear transverse waves in shallow water, ion-acoustic waves in plasma, elastic media, optical fibres, hydromagnetic wave in cold plasma [11]. The development of an equal width undular bore is investigated in [12] and compared with that of the RLW bore. The modified equal width (MEW) equation

    ut+a(u3)xbuxxt=0, (1.3)

    is based upon the EW (1.1). (1.3) is a nonlinear wave equation with cubic nonlinearity which admits solitary wave solutions with the same width, which is considered to explain many physical phenomena [13,14,15].

    Based on many applications of MEW equation, there is a variety of investigations on MEW equation. In 2000, the author studied the solitary wave motion and interaction for MEW equation by finite element methods [16]. The modified equal width equation and its variants were investigated by Wazwaz [8] by a sine-cosine ansatz and the tanh method, compactons, solitons, solitary patterns, and periodic solution are obtained. In [17], the propagation of the solitary wave for time split MEW equation and space split MEW equation was investigated by quintic B-spline collocation method. After that, the numerical solution of the MEW equation was proposed by the collocation method using the radial basis functions with first order accurate forward difference approximation [18]. Lu [19] solved the modified equal width equation by variational iteration method which provides remarkable accuracy in comparison with the analytical solution. Cheng and Liew [20] derived formulae for an improved element-free Galerkin method for MEW equation by numerical examples. In [21], the motion of a single solitary wave and interaction of two solitary waves for the MEW equation were studied. More recently, Shi and Zhang [22] obtained the periodic solutions, dark solutions, soliton solutions and soliton-like solutions of the space-time fractional MEW equation by ansatz method. Additionally, there are lots of investigations on finding the traveling wave solutions for combined systems based on MEW equation, such as, ZK-MEW equation [23,24], KP-MEW equation [25,26], KP-MEW-Burgers [27]. However, there is few investigation on the traveling wave solution especially periodic wave solution for the perturbed MEW equation.

    In this paper, we consider the following perturbed MEW equation

    ut+(u2)x+uxxt+ε((uux)xuxx+uxxxx)=0, (1.4)

    where ε is a small nonnegative parameter. In Eq (1.4), ut is the evolution term, the second term represents the nonlinear term, uxx is the backward diffusion, uxxxx is the dissipative term, (uux)x is the nonlinear term when the Marangoni effect is considered which describing the opposite to the Benard convection [28,29]. The Eq (1.4) describes the water motion in a wide range of weak dissipative circumstances. We focus on studying the existence of the periodic wave solution for perturbed MEW equation (1.4) and finding the number of periodic wave solution by using geometric singular perturbation theory. Furthermore, for the persisted periodic wave solution, we give the monotonicity of the period.

    On the topic about finding the existence of traveling wave solution for perturbed nonlinear wave equations by using geometric singular perturbation theory, there are also lots of excellent productions. Ogama [30] established the existence of solitary waves and periodic waves for the perturbed KdV equation. Fan and Tian [31] studied the existence of solitary wave solution of perturbed mKdV-KS equation. Tang [32] gave the condition of solitary wave solution persisted. Mansour [33] constructed solitary waves for a generalized nonlinear dispersive-dissipative equation. Yan et al. [34] proved the persistence of solitary waves and periodic waves to a perturbed generalized KdV equation. By investigating the ratio of Abelian integrals, Du and his cooperators proved the existence of traveling wave solutions for some delayed nonlinear wave equations [35,36,37,38]. Chen [39,40] studied the existence of traveling wave solutions for perturbed KdV equation. In references [41,42], two different generalized perturbed BBM equations were considered. Motivated by the references, we present the existence of uniqueness isolated periodic wave solution of (1.4) by using geometric singular perturbation theory. Combing with the Chebyshev system criterion and symbolic computation, the monotonicity of the ratio of Abelian integrals is given. Moreover, for the periodic wave solution, the property of the wave speed, the monotonicity and the range of period are obtained.

    The rest of this paper is organized as follows. In Section 2, the geometric singular perturbation theory is introduced and our main result is stated. In Section 3, the existence of unique periodic wave for the perturbed MEW equation is proved by using geometric singular perturbation theory. Chebyshev system and symbolic computation are used to verify the monotonicity of the ratio of Abelian integrals, which is more effective than the method used in the references [39,40,41]. In Section 4, we investigate the monotonicity of period by Picard-Fuchs equation, which is not been considered in the references [42,43].

    We aim to prove the traveling wave persists for sufficiently small ε>0 by using geometric singular perturbation theory, so we introduce the geometric singular perturbation theory which is due to [44,45], firstly.

    Consider the system

    {x(t)=f(x,y,ε),y(t)=εg(x,y,ε), (2.1)

    where =ddt, 0<ε1 is a real and small parameter, x=(x1,x2,,xk)TRk, y=(y1,y2,,yl)TRl. The following hypothesis about the system (2.1) is needed.

    (H1) The functions f and g are both Cr on a set U×I, where 0<r<+ URk+l is open and I is an open interval containing 0.

    With a change of time scaling τ=εt, system (2.1) can be written as

    {ε˙x=f(x,y,ε),˙y=g(x,y,ε), (2.2)

    where ˙=ddτ. The time scale τ is slow and t is fast. When ε0, systems (2.1) and (2.2) are equivalent. Thus system (2.1) is called the fast system, while system (2.2) is called the slow system. In (2.1), letting ε0, we obtain the layer system

    {x(t)=f(x,y,0),y(t)=0. (2.3)

    Here, x is called the fast variable, whereas y is called the slow variable. Let ε0 in (2.2), the limit only makes sense if f(x,y,0)=0 and is given by

    {f(x,y,0)=0,˙y=g(x,y,0). (2.4)

    We shall assume that we are given an l-dimensional manifold, possibly with boundary, M0 which is contained in the set {f(x,y,0)=0}. The manifold M0 is normally hyperbolic if the layer system (2.3) at each point in M0 has exactly l eigenvalues on the imaginary axis. Moreover, we give the following hypothesis.

    (H2) The set M0 is a compact manifold, possibly with boundary, and is normally hyperbolic relative to (2.3).

    Definition 2.1. A set M is locally invariant under the flow from (2.1) if it has neighborhood V so that no trajectory can leave M without also leaving V. In other words, it is locally invariant if for all xM, x[0,t]V implies that x[0,t]M, similarly with [0,t] replaced by [t,0], when t<0, where the notation xt is used to denote the application of a flow after time t to the initial condition x.

    Assume that there is a Cr function h0(y), 0<r<+, with K being a compact domain in Rl, such that M0={(x,y):x=h0(y)}. Consequently, the following geometric theory of singular perturbation is established in [44].

    Lemma 2.1. For ε>0 is sufficiently small, there exists a manifold Mε lying within O(ε) of M0. Mε is diffeomorphic to M0 and locally invariant under the flow of (2.1), and Cr in x,y and ε, for any 0<r<+.

    Lemma 2.2. Under the hypotheses (H1) and (H2), for ε>0 is sufficiently small, there exists a function x=hε(y) defining on K, such that the graph

    Mε={(x,y):x=hε(y)},

    is locally invariant under (2.1). Moreover, hε(y) is Cr, for any 0<r<+, jointly in y and ε. Mε possesses locally invariant stable and unstable manifold Ws(Mε) and Wu(Mε) lying within O(ε) and being Cr diffeomorphic to the stable and unstable manifold Ws(M0) and Wu(M0) of the critical manifold M0.

    For Eq (1.4), making the traveling wave transformation u(x,t)=u(x+ct)=u(ξ), where c is the wave speed. Integrating it once and neglecting the integral constant, then corresponding traveling wave system is

    cu+u2+cu+ε(uuu+u)=0, (2.5)

    where is the derivative respect to ξ. Taking a time scale transformation u=cϕ and ξ=z to (2.5), it obtains

    ϕ+ϕ2+d2ϕdz2+ε(ϕϕ1cdϕdz+1cd3ϕdz3)=0, (2.6)

    which is equivalent to the following three-dimensional system

    {dϕdz=y,dydz=ω,ε1cdωdz=ϕϕ2ωε(yc+ϕy). (2.7)

    When ε=0, system (2.7) corresponds to a unperturbed system

    {dϕdz=y,dydz=ϕϕ2, (2.8)

    which is a Hamiltonian system with the energy function

    H(ϕ,y)=y22+ϕ22+ϕ33. (2.9)

    Clearly, (2.8) has two equilibrium points (1,0) and (0,0). The origin (0,0) is a center and (1,0) is a saddle. It is well known that (2.8) is determined by its potential energy function and its equilibrium points. H(1,0)=H(12,0)=16, H(0,0)=0. Figure 1 shows a family of closed orbits surrounded by a homoclinic loop.

    Figure 1.  The portrait of system (2.8).

    Let Γh be the closed orbits defined by H(ϕ,y)=h which surrounds a center. Suppose that ϕ(z,h) is a point lying the corresponding closed orbit Γh, ϕ(z,ε,h,c(ε,h)) is the periodic wave solution of (2.5) near Γh on c=c(ε,h). Denote that T(h) is the period of ϕ(z,ε,h,c(ε,h)), ϕ0(z) is the solution corresponds to homoclinic loop, then for the traveling wave system of perturbed MEW equation (1.4), we obtain the following statements.

    Theorem 2.1. For any given traveling wave speed c(12,+), there exists ε0(c)>0 such that when 0<ε<ε0(c), then Eq (1.4) has a unique isolated periodic wave solution u(x+ct), which is given by u(x+ct)=cϕ(x+ct,c,ε,h(c,ε)), satisfying

    limε0ϕ(x+ct,c,ε,h)=ϕ(x+ct,h),lim(c,ε)(+,0),0<ε<ε0(c)ϕ(x+ct,c,ε,h)0,lim(c,ε)(12,0),0<ε<ε0(c)ϕ(x+ct,c,ε,h)ϕ0(x+ct).

    Furthermore, c(ε,h) satisfies

    limε0c(ε,h)=c(h),c(τ,h)h<0,

    where c(h) is a strictly decreasing function in h satisfying 12<c(h)<+.

    Theorem 2.2. The period of the isolated periodic wave solution shown in Theorem 2.1 is strictly increasing for h(0,16) and satisfies

    limh1/6T(h)=+,limh0T(h)=2π.

    By the geometric singular perturbation theory, when ε0, the transformation z=ετ is introduced to change the system (2.7) to

    {dϕdτ=εy,dydτ=εω,1cdωdτ=ϕϕ2ωε(1cy+ϕy). (3.1)

    System (2.7) is the slow system and (3.1) is the fast system, they are equivalent when ε>0. The two different time-scales corresponds to two different limiting systems. If ε=0, the flow of system (3.1) is confined to the two-dimensional invariant manifold

    M0={(ϕ,y,ω)R3:ω=ϕϕ2}

    and its dynamics are determined only by the first two equations of (3.1). The set M0 is the slow manifold. Since the linearized matrix of (3.1) with ε=0 is

    (000000c(1+2ϕ)0c),

    It is not difficult to find that the eigenvalues are 0, 0, c, then the slow manifold M0 is normally hyperbolic. From Lemma 2.2, there exists a sub-manifold Mε of the perturbed system (2.7) of R3 for sufficiently small ε>0, which can be written as

    Mε={(ϕ,y,ω)R3:ω=ϕϕ2+g(ϕ,y,ε)},

    where g is a smooth function defined on a compact domain and satisfies g(ϕ,y,0)=0. Then assume that g(ϕ,y,ε) is expanded into Taylor series g(ϕ,y,ε)=εg1(ϕ,y)+O(ε2). Substituting ω=ϕϕ2+g(ϕ,y,ε) into the slow system (2.7) and comparing the coefficient of ε, it obtains

    g1=2cy+(2c1)ϕy.

    Therefore, the slow system (2.7) restricted on Mε is given by a regular perturbed system

    {dϕdz=y,dydz=ϕϕ2+ε(2cy+2ccϕy)+O(ε2). (3.2)

    For h(0,1/6), suppose that there exists a closed orbit Γh of (3.2)|ε=0 which surrounds the center (0,0). T(h) is the period of Γh. A(h)Γh is the rightmost point on the positive ϕ-axis at z=0. For ε>0 sufficiently small, let Γhε be a piece of the orbit for (3.2) starting from A(h) to the next intersection point B(hε) with the positive ϕ-axis at z=z(ε) for 0<|hεh|1. By [46], the displacement function between B(hε) and A(h) is given by

    d(h,c,ε)=^ABdH=^AB(ϕ+ϕ2)dϕ+ydy=z(ε)0{(ϕ+ϕ2)y+[ϕϕ2+ε(2cy+2ccϕy)+O(ε2)]y}dz=ε1cz(ε)0[2y2+(2c)ϕy2+O(ε)]dzεΦ(h,c,ε).

    By continuousness theorem, we have

    limε0Γh,ε=Γh,limε0B(h,ε)=B(h),limε0z(ε)=T(h).

    Therefore, it has

    Φ(h,c,ε)=1cM(h,c)+O(ε),

    where

    M(h,c)=Γh[2y2+(2c)ϕy2]dz=Γh[2y+(2c)ϕy]dϕ=2J0(h)+(2c)J1(h), (3.3)

    with Ji(h)=Γhϕiydϕ, i = 0, 1. M(h,c) is called Melnikov function. By the Poincaré bifurcation theory, the isolated zeros of d(h,ε) corresponds to limit cycles of system (3.2).

    In order to investigate the zero of M(h,c), the following lemmas are needed.

    Lemma 3.1. For h(0,16), J0(h)>0 and J0(h)>0.

    Proof. Denote (ϕ,y) is a point lying on Γh. By Green formula, it yields

    J0(h)=Γhydϕ=intΓhdϕdy>0.

    Since y2=2hϕ223ϕ3, it has Ji(h)=Γhϕiyhdϕ=Γhϕiydϕ, then

    J0(h)=Γh1ydϕ=T(h)0dz=T(h)>0.

    Denote that P(h)=J1(h)J0(h), then

    M(h,c)=J0[2+(2c)P(h)]. (3.4)

    Lemma 3.2. J0(16)=65,J1(16)=635. Then, J1(16)J0(16)=17.

    Proof. Denote that α(h) and β(h) are the left and right intersection points of Γh to ϕ-axis, respectively. From (2.9), it has y=±23ϕ3ϕ2+2h and α(16)=1,β(16)=12, then we compute the J0(16) and J1(16) directly

    J0(16)=Γhydϕ=212123ϕ3ϕ2+13dϕ=263121(1+ϕ)12ϕdϕ=65,J1(16)=Γhϕydϕ=2121ϕ23ϕ3ϕ2+13dϕ=263121ϕ(1+ϕ)12ϕdϕ=635,

    then it obtains the statements.

    Lemma 3.3. limh0J1(h)J0(h)=0.

    Proof. When h0, Γh approaches to the center (0,0), implying that ϕ0. By applying mean value theorem for integrals, it has

    limh0J1(0)J0(0)=limh0ΓhϕydϕΓhydϕ=limϕ0ϕ=0.

    Lemma 3.4. For h(0,16), J1(h)J0(h) is decreasing strictly from 0 to 17.

    Proof. Monotonicity of J1(h)J0(h) on (0,16) is equivalent to {J0(h),J1(h)} is extended complete Chebyshev system [47], i.e. any nontrivial linear combination a0J0(h)+a2J1(h) has at most one zero on (0,16). Denote f0(ϕ)=1, f1(ϕ)=ϕ. Setting that

    li(ϕ)=(fiΨ)(ϕ)(fiΨ)(z(ϕ)),

    where Ψ(ϕ)=12ϕ2+13ϕ3 and z(ϕ) is an involution function: (1,0)(0,12) by Ψ(ϕ)=Ψ(z(ϕ)). Since Ψ(ϕ)Ψ(z(ϕ))=16(ϕz)q(ϕ,z), where q(ϕ,y)=2ϕ2+2ϕz+2z2+3ϕ+3z. We need to prove two Wronskians W[l0(ϕ)], W[l0(ϕ),l1(ϕ)] are non-vanishing on (1,0). With aids of Maple, it has

    W[l0(ϕ)]=(ϕz)(ϕ+z+1)ϕz(ϕ+1)(z+1),W[l0(ϕ),l1(ϕ)]=(ϕz)3w0(ϕ,z)ϕ2z2(ϕ+1)2(z+1)2(2ϕ+4z+3),

    where w0(ϕ,z)=4ϕ2+6ϕz+4z2+7ϕ+7z+3. It is not hard to verify that for ϕ(1,0),z(0,1/2), it has ϕ+z+10, then W[l0(ϕ)]0. In order to check the zero of W[l0(ϕ),l1(ϕ)], we calculate the resultants of 2ϕ+4z+3 and q(ϕ,z), w0(ϕ,z) and q(ϕ,z) respect to z. It has

    R(2ϕ+4z+3,q,z)=6(2ϕ+3)(2ϕ1)0,R(w0(ϕ,z),q,z)=2(8ϕ4+16ϕ34ϕ212ϕ+9)0

    for ϕ(0,1), then W[l0(ϕ),l1(ϕ)]0. Therefore, J1(h)J0(h) is monotonic on (0,16). By Lemmas 3.2 and (3.3), the assertion in this lemma is proved.

    Remark 3.1. Here, the Chebyshev system criterion is used to prove the monotonicity of the ratio of Abelian integrals, which is more effective and simpler then linear programming method used in the references [39,40,41].

    From (3.4), we know that for each h(0,16), when c=c(h)=(2+2P(h))/P(h), it derives M(h,c)=0. Moreover, the monotonicity of P(h) implies that the zero of P(h) is unique, denoted by h, and c(h)=2/P(h)2<0, then from Lemma 3.4, we get

    12<c(h)<,limh16c(h)=12,limh0c(h)=+.

    Combining with implicit function theorem, there exists c=c(h)+O(ε) such that M(h,c)+O(ε) has a unique zero near h. Therefore, the conclusion in Theorem 2.1 is obtained.

    For the sections before, our purpose is focusing on the analysis of the relationship between the speed of periodic wave solutions and the level h. Moreover, the property of the period is also significant in reality. On this purpose, it is needed to introduce some additional properties for Ji(h) and Ji(h), i=0,1,2.

    From (2.9), we have

    Ji(h)=Γhϕiydϕ=2βαϕi23ϕ3ϕ2+2hdϕ=2βαϕiE(ϕ)dϕ,

    where E(ϕ)=23ϕ3ϕ2+2h, h(0,16). The derivative of the Abelian integrals is

    Ji(h)=Γhϕiyhdϕ=Γhϕiydϕ=2βαϕiEdϕ.

    Then T(h) is a period of ϕ(z) satisfying T(h)=2βαdz=Γh1ydϕ=J0(h). Therefore, we have the following lemmas.

    Lemma 4.1. (J0J1)=Λ(h)(J0J1), where Λ(h)=135(42h76h30h6).

    Proof. Since EdEdϕ=ϕ2ϕ, it has

    J0(h)=2βαEdϕ=2βαE2dϕE=2βα(23ϕ3ϕ2+2h)dϕE=2βα23ϕ(EdEdϕ+ϕ)dϕE2βαϕ2dϕE+2hΓhdϕE=43βαϕdE23βαϕ2dϕE+4hβαdϕE.

    From integration by part, βαϕdE=ϕE|βαβαEdϕ=12J0. Moreover,

    βαϕ2dϕE=βα(EdEdϕϕ)dϕE=βαdEβαϕdϕE=12J1.

    Thus,

    J0=23J0+13J1+2hJ0,

    therefore, it obtains

    J0=15(6hJ0+J1).

    Similarly, we get

    J1(h)=2βαϕEdϕ=2βαϕE2dϕE=2βαϕ(23ϕ3ϕ2+2h)dϕE=βα43ϕ2(EdEdϕ+ϕ)dϕE2βαϕ3dϕE+4hβαϕdϕE=43βαϕ2dE23βαϕ3dϕE+4hβαϕdϕE. (4.1)

    Moreover, since

    βαϕ2dE=ϕ2E|βα2βαϕEdϕ=J1,βαϕ3dϕE=βαϕ(EdEdϕϕ)dϕE=βαϕdEβαϕ2dϕE=12J0+12J1, (4.2)

    substitute (4.2) into (4.1), it obtains

    J1=635[hJ0+(5h1)J1].

    Therefore the proof of the lemma is completed. From Lemma 4.1, we obtain the following lemmas.

    Lemma 4.2. J0 and J1 satisfy the Picard-Fuchs equation

    (J0J1)=1h(6h1)(5h176h7h)(J0J1).

    Lemma 4.3. J0 and J1 satisfy the Picard-Fuchs equation

    (J0J1)=1h(6h1)(h16hh)(J0J1).

    Proof. Denote that J=(J0,J1)T. From lemmas 7 and 8, we have J=Λ1(IΛ)J, where I is the unit matrix, Λ is Λ(h) derivative with respect to h. It is not hard to verify that

    Λ1(IΛ)=1h(6h1)(5h176hh)(15063517)=1h(6h1)(h16hh).

    This proves the lemma.

    Lemma 4.4. limh1/6T(h)=+,limh0T(h)=2π. Furthermore, For h(0,1/6), T(h)>0.

    Proof. Since h=1/6 corresponds the homoclinic loop connecting to saddle (1,0), it is not hard to verify the first statement. For the latter, by Lemma 4.2, it obtains

    limh0T(h)=limh0J0(h)=limh01h(6h1)[(5h1)J076J1]=limh0J0h(6h1)[(5h1)76J1J0]=limh0(5h1)J0h(6h1)=56limh0J0h=2π.

    From Lemmas 4.2 and 4.3, it obtains

    T(h)=J0(h)=1h(6h1)[hJ016J1]=1h2(6h1)2[h((5h1)J076J1)16(hJ0+7hJ1)]=5h(6h1)2(h16)J0>0.

    Therefore we get the Theorem 2.2.

    Remark 4.1. In this section, we show the properties of period for the uniqueness isolated periodic wave solution, which were not considered in the references [42,43].

    This paper mainly proves the existence of unique periodic wave solution for perturbed MEW equation with weak backward diffusion, dissipation and Marangoni effect. By geometric singular perturbation theory, the local invariant submainfold is given, and then the singular perturbation is reduced into regular perturbation. We established the existence of periodic wave on for perturbed MEW equation by analyzing the monotonicity of the ratio of Abelian integrals. Chebyshev system criterion is utilized to prove the uniqueness of the periodic wave solution. Particularly, the related properties on the periodic wave are given by Picard-Fuchs equation.

    The authors would like to thank the anonymous reviewers for providing useful comments and suggestions which help to strengthen the manuscript. The author is supported by Guangxi First-class Discipline statistics Construction Project Fund (No. 2022SXYB10).

    This work does not have any conflict of interest.



    [1] T. M. Apostol, Arithmetical properties of generalized Ramanujan Sums, Pacific J. Math., 41 (1972), 281-293. doi: 10.2140/pjm.1972.41.281
    [2] T. M. Apostol, Introduction to analytic number theory, Springer-Verlag, New York, 1976.
    [3] S. Beslin and S. Ligh, Another generalisation of Smith's determinant, Bull. Austral. Math. Soc., 40 (1989), 413-415. doi: 10.1017/S0004972700017457
    [4] R. Bhatia, J. A. Dias da Silva, Infinite divisibility of GCD matrices, Amer. Math. Monthly, 115 (2008), 551-553. doi: 10.1080/00029890.2008.11920562
    [5] K. Bourque and S. Ligh, On GCD and LCM matrices, Linear Algebra Appl., 174 (1992), 65-74. doi: 10.1016/0024-3795(92)90042-9
    [6] K. Bourque and S. Ligh, Matrices associated with arithmetical functions, Linear Multilinear Algebra, 34 (1993), 261-267. doi: 10.1080/03081089308818225
    [7] K. Bourque and S. Ligh, Matrices associated with classes of arithmetical functions, J. Number Theory, 45 (1993), 367-376. doi: 10.1006/jnth.1993.1083
    [8] K. Bourque and S. Ligh, Matrices associated with classes of multiplicative functions, Linear Algebra Appl., 216 (1995), 267-275. doi: 10.1016/0024-3795(93)00154-R
    [9] W. Cao, On Hong's conjecture for power LCM matrices, Czechoslovak Math. J., 57 (2007), 253-268. doi: 10.1007/s10587-007-0059-3
    [10] P. Codeca and M. Nair, Calculating a determinant associated with multilplicative functions, Boll. Unione Mat. Ital. Sez. B Artic. Ric. Mat., 8 (2002), 545-555.
    [11] W. D. Feng, S. F. Hong and J. R. Zhao, Divisibility properties of power LCM matrices by power GCD matrices on gcd-closed sets, Discrete Math., 309 (2009), 2627-2639. doi: 10.1016/j.disc.2008.06.014
    [12] P. Haukkanen and I. Korkee, Notes on the divisibility of GCD and LCM matrices, Int. J. Math. Math. Sci., (2005), 925-935.
    [13] T. Hilberdink, Determinants of multiplicative Toeplitz matrices, Acta Arith., 125 (2006), 265-284. doi: 10.4064/aa125-3-4
    [14] S. A. Hong, S. N. Hu and Z. B. Lin, On a certain arithmetical determinant, Acta Math. Hungar., 150 (2016), 372-382. doi: 10.1007/s10474-016-0664-4
    [15] S. A. Hong and Z. B. Lin, New results on the value of a certain arithmetical determinant, Publicationes Mathematicae Debrecen, 93 (2018), 171-187. doi: 10.5486/PMD.2018.8139
    [16] S. F. Hong, On Bourque-Ligh conjecture of LCM matrices, Adv. Math. (China), 25 (1996), 566-568.
    [17] S. F. Hong, On the Bourque-Ligh conjecture of least common multiple matrices, J. Algebra, 218 (1999), 216-228. doi: 10.1006/jabr.1998.7844
    [18] S. F. Hong, Gcd-closed sets and determinants of matrices associated with arithmetical functions, Acta Arith., 101 (2002), 321-332. doi: 10.4064/aa101-4-2
    [19] S. F. Hong, On the factorization of LCM matrices on gcd-closed sets, Linear Algebra Appl., 345 (2002), 225-233. doi: 10.1016/S0024-3795(01)00499-2
    [20] S. F. Hong, Factorization of matrices associated with classes of arithmetical functions, Colloq. Math., 98 (2003), 113-123. doi: 10.4064/cm98-1-9
    [21] S. F. Hong, Nonsingularity of matrices associated with classes of arithmetical functions, J. Algebra, 281 (2004), 1-14. doi: 10.1016/j.jalgebra.2004.07.026
    [22] S. F. Hong, Divisibility properties of power GCD matrices and power LCM matrices, Linear Algebra Appl., 428 (2008), 1001-1008. doi: 10.1016/j.laa.2007.08.037
    [23] S. F. Hong and K. S. Enoch Lee, Asymptotic behavior of eigenvalues of reciprocal power LCM matrices, Glasgow Math. J., 50 (2008), 163-174. doi: 10.1017/S0017089507003953
    [24] S. F. Hong and R. Loewy, Asymptotic behavior of eigenvalues of greatest common divisor matrices, Glasgow Math. J., 46 (2004), 551-569. doi: 10.1017/S0017089504001995
    [25] S. F. Hong and R. Loewy, Asymptotic behavior of the smallest eigenvalue of matrices associated with completely even functions (mod r), Int. J. Number Theory, 7 (2011), 1681-1704. doi: 10.1142/S179304211100437X
    [26] S. F. Hong, J. R. Zhao and Y. Z. Yin, Divisibility properties of Smith matrices, Acta Arith., 132 (2008), 161-175. doi: 10.4064/aa132-2-4
    [27] I. Korkee and P. Haukkanen, On the divisibility of meet and join matrices, Linear Algebra Appl., 429 (2008), 1929-1943. doi: 10.1016/j.laa.2008.05.025
    [28] M. Li, Notes on Hong's conjectures of real number power LCM matrices, J. Algebra, 315 (2007), 654-664. doi: 10.1016/j.jalgebra.2007.05.005
    [29] M. Li and Q. R. Tan, Divisibility of matrices associated with multiplicative functions, Discrete Math., 311 (2011), 2276-2282. doi: 10.1016/j.disc.2011.07.015
    [30] Z. B. Lin and S. A. Hong, More on a certain arithmetical determinant, Bull. Aust. Math. Soc., 97 (2018), 15-25. doi: 10.1017/S0004972717000788
    [31] M. Mattila, On the eigenvalues of combined meet and join matrices, Linear Algebra Appl., 466 (2015), 1-20. doi: 10.1016/j.laa.2014.10.001
    [32] P. J. McCarthy, A generalization of Smith's determinant, Can. Math. Bull., 29 (1986), 109-113. doi: 10.4153/CMB-1986-020-1
    [33] H. J. S. Smith, On the value of a certain arithmetical determinant, Proc. London Math. Soc., 7 (1875), 208-213.
    [34] Q. R. Tan, Divisibility among power GCD matrices and among power LCM matrices on two coprime divisor chains, Linear Multilinear Algebra, 58 (2010), 659-671. doi: 10.1080/03081080903008579
    [35] Q. R. Tan and M. Li, Divisibility among power GCD matrices and among power LCM matrices on finitely many coprime divisor chains, Linear Algebra Appl., 438 (2013), 1454-1466. doi: 10.1016/j.laa.2012.08.036
    [36] Q. R. Tan, Z. B. Lin and L. Liu, Divisibility among power GCD matrices and among power LCM matrices on two coprime divisor chains II, Linear Multilinear Algebra, 59 (2011), 969-983. doi: 10.1080/03081087.2010.509721
    [37] Q. R. Tan, M. Luo and Z. B. Lin, Determinants and divisibility of power GCD and power LCM matrices on finitely many coprime divisor chains, Appl. Math. Comput., 219 (2013), 8112-8120.
    [38] J. X. Wan, S. N. Hu and Q. R. Tan, New results on nonsingular power LCM matrices, Electron. J. Linear Al., 27 (2014), 258.
    [39] A. Wintner, Diophantine approximations and Hilbert's space, Am. J. Math., 66 (1944), 564.
    [40] J. Xu and M. Li, Divisibility among power GCD matrices and among power LCM matrices on three coprime divisor chains, Linear Multilinear Algebra, 59 (2011), 773-788. doi: 10.1080/03081087.2010.526942
    [41] Y. Yamasaki, Arithmetical properties of multiple Ramanujan sums, Ramanujan J., 21 (2010), 241-261. doi: 10.1007/s11139-010-9223-8
    [42] J. R. Zhao, Divisibility of power LCM matrices by power GCD matrices on gcd-closed sets, Linear Multilinear Algebra, 62 (2014), 735-748. doi: 10.1080/03081087.2013.786717
  • Reader Comments
  • © 2020 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(4131) PDF downloads(398) Cited by(3)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog