Research article

On oscillatory second order impulsive neutral difference equations

  • Received: 12 October 2019 Accepted: 01 March 2020 Published: 06 March 2020
  • MSC : 39A10, 39A12, 39A21

  • The present paper deals with the problem of oscillation for a class of second order nonlinear neutral impulsive difference equations with fixed moments of impulse effect. The technique employed here is due to the classical impulsive inequalities. Some examples are given to illustrate our results.

    Citation: Gokula Nanda Chhatria. On oscillatory second order impulsive neutral difference equations[J]. AIMS Mathematics, 2020, 5(3): 2433-2447. doi: 10.3934/math.2020161

    Related Papers:

    [1] Muhammad Tariq, Sotiris K. Ntouyas, Hijaz Ahmad, Asif Ali Shaikh, Bandar Almohsen, Evren Hincal . A comprehensive review of Grüss-type fractional integral inequality. AIMS Mathematics, 2024, 9(1): 2244-2281. doi: 10.3934/math.2024112
    [2] Abd-Allah Hyder, Mohamed A. Barakat, Doaa Rizk, Rasool Shah, Kamsing Nonlaopon . Study of HIV model via recent improved fractional differential and integral operators. AIMS Mathematics, 2023, 8(1): 1656-1671. doi: 10.3934/math.2023084
    [3] Mustafa Gürbüz, Yakup Taşdan, Erhan Set . Ostrowski type inequalities via the Katugampola fractional integrals. AIMS Mathematics, 2020, 5(1): 42-53. doi: 10.3934/math.2020004
    [4] Jun Moon . The Pontryagin type maximum principle for Caputo fractional optimal control problems with terminal and running state constraints. AIMS Mathematics, 2025, 10(1): 884-920. doi: 10.3934/math.2025042
    [5] Ismail Gad Ameen, Dumitru Baleanu, Hussien Shafei Hussien . Efficient method for solving nonlinear weakly singular kernel fractional integro-differential equations. AIMS Mathematics, 2024, 9(6): 15819-15836. doi: 10.3934/math.2024764
    [6] Pinghua Yang, Caixia Yang . The new general solution for a class of fractional-order impulsive differential equations involving the Riemann-Liouville type Hadamard fractional derivative. AIMS Mathematics, 2023, 8(5): 11837-11850. doi: 10.3934/math.2023599
    [7] Iyad Suwan, Mohammed S. Abdo, Thabet Abdeljawad, Mohammed M. Matar, Abdellatif Boutiara, Mohammed A. Almalahi . Existence theorems for Ψ-fractional hybrid systems with periodic boundary conditions. AIMS Mathematics, 2022, 7(1): 171-186. doi: 10.3934/math.2022010
    [8] Najat Almutairi, Sayed Saber . Chaos control and numerical solution of time-varying fractional Newton-Leipnik system using fractional Atangana-Baleanu derivatives. AIMS Mathematics, 2023, 8(11): 25863-25887. doi: 10.3934/math.20231319
    [9] Wei Zhang, Jifeng Zhang, Jinbo Ni . New Lyapunov-type inequalities for fractional multi-point boundary value problems involving Hilfer-Katugampola fractional derivative. AIMS Mathematics, 2022, 7(1): 1074-1094. doi: 10.3934/math.2022064
    [10] Muneerah AL Nuwairan, Ahmed Gamal Ibrahim . The weighted generalized Atangana-Baleanu fractional derivative in banach spaces- definition and applications. AIMS Mathematics, 2024, 9(12): 36293-36335. doi: 10.3934/math.20241722
  • The present paper deals with the problem of oscillation for a class of second order nonlinear neutral impulsive difference equations with fixed moments of impulse effect. The technique employed here is due to the classical impulsive inequalities. Some examples are given to illustrate our results.


    The inverse problem for differential equations is part of the fascinating branches of mathematics. It is developed in connection with diverse branches of applied science [1]. The type of boundary condition determines the identity of the problem. If the boundary is given for an initial time the problem is recognized as an initial value problem (IVP); if the boundary is described for the final, it is called a terminal value problem (TVP). If the boundary is described at both times, it is a Sturm-Liouville problem [2].

    For ordinary differential equations (ODEs), the terminal value problem can be transformed into an initial value problem, and they are generally well-posed problems. Surprisingly, this is not true for fractional differential equations (FDEs). Not only can a TVP for FDEs not be transformed into an IVP, but also such problems are not well-posed in general [3,4]. It has been recently noticed that a TVP for FDEs is well-posed in a small neighborhood of the boundary [5,6,7,8,9].

    In recent decades, literature on fractional operators has been extensively increased for describing non-local dynamical processes [10]. The non-local property of such processes is usually captured by an integral with a memory kernel. Recently, in the notable paper [11], high dimensional problems are reduced to an integral equation with a memory kernel. This is a green light for replacing ordinary derivatives with fractional derivatives in complex processes. Thus, fractional differential equations can describe abundant models well.

    There is no unified definition for a fractional derivative. However, if we restrict some properties (such as being singular or non-singular, local or non-local) we can obtain some completely separate classifications. Among them, the Caputo derivative has received more interest and attention in the literature, mainly for well-interpreted boundary conditions and more similarity to the ordinary derivative (the Caputo derivative of the constant function is zero) [12].

    Similarly, fractional integrals have diverse definitions. The route toward the generalized fractional operator will be clear if we monitor some selected classes of fractional integrals. The Riemann-Liouville fractional integral is a generalization of the integer order integral operator

    Inf(x)=xadt1t1adt2tn1af(tn)dtn=1(n1)!xa(xt)n1f(t)dt,

    and the Hadamard fractional integral is a generalization of the nth order integral operator

    HInf(x)=xadt1t1t1adt2t2tn1af(tn)dtntn              =1(n1)!xa(ln(xt))n1f(t)dtt

    where nN [13]. Katugampola [14,15] introduced his fractional derivative by replacing ti with tρi, ρ>0. Fu et al. have replaced exponential functions to introduce exponential fractional derivatives [16]. It is tempting to replace ti with a general function g(ti), (where g is a weighted function) to study generalized fractional derivatives. Such unified generalization of fractional operators with respect to another function has been studied in [17,18].

    It is a common aspect of pure mathematics to see an application of a seemingly not applicable concept later. Analogously, for generalized fractional derivatives, we see some noticeable applications that have been published just very recently. For example, the subdiffusion equation with a generalized fractional derivative has been effectively invoked to describe subdiffusion in a medium with an evolving structure over time [19]. The application of generalized fractional operators for the Fokker-Planck equation has been noticed in [16,20]. We will show another importance of such generalization in inverse problems.

    Considering the generality of the fractional derivative with respect to another derivative and its application, it is highly motivated to study fractional differential equations with respect to another derivative.

    A ν-dimensional terminal value problem for a system of generalized fractional differential equations (GFDEs) is described by

    GaDˉα,gty=f(t,y), t[a,b] (1.2)

    and

    y(b)=yb, (1.3)

    where GaDˉα,gt is a generalized fractional derivative with respect to the vector function g=[g1,,gν]T (gi:[a,b]R are strictly increasing functions with continuous derivatives gi on (a,b)), ˉα=[α1,,αν]T (αi(0,1)), y:[a,b]Rn is an unknown vector function, f:[a,b]×RnRn is a given vector function describing an evolutionary processes, and ybRn is a terminal value.

    In a real dynamical process, we do not expect each separate component to have the same memory. Therefore, the study of FDEs with vector order is more important than that for fixed order. We note that the vector order introduced in [21] is an element-wise operator and defined by

    GaDˉα,gty=[GaDα1,g1ty1,,GaDαν,gνtyν].

    This paper contributes the following achievements:

    Study of the TVP for systems of high dimensional nonlinear generalized FDEs,

    Computing generalizing fractional operators by classical fractional operators,

    Introducing suitable weighted space that relates previous studies to generalized fractional derivative,

    Introducing and analyzing a comprehensive collocation method that covers all generalized operators,

    Obtaining a lower bound for well-posedness of TVPs with various weight functions,

    Applying generalized derivatives for estimating the past population of diabetes cases.

    To the best of our knowledge, these are investigated for the first time. The important achievement of this paper is that it gives us more analyzed options in modeling dynamical processes, to choose a better weight function. If we need an inverse problem with a longer interval, the result of this paper introduces a generalized derivative that guarantees the well-posedness in a study model.

    A generalized fractional derivative can be transformed into a classical fractional derivative [13,22]. This fact was noticed and employed for solving generalized fractional differential equations, especially in [22]. In this respect, we introduced a weighted space with respect to another function to carry out our analysis from generalized fractional-order derivatives toward classical fractional derivatives.

    Numerical methods in parallel to theoretical analysis for solving fractional differential equations have developed rapidly. The Jacobi spectral methods for solving related functional integral and fractional differential equations have been extensively studied in [23,24,25,26]. Applying this method for terminal value problems can be found in [27]. The collocation methods on piecewise polynomials spaces for fractional differential equations were studied in [6,9]. The spectral methods provide fast convergence methods that transform a related problem into a high-dimensional algebraic equation, while the collocation method provides a high order method with more options for control of the error. As far as we know, the collocation method for terminal value problems with generalized fractional differential equations is not yet studied and not analyzed. An important aim of this paper also is to propose and analyze such methods for these problems.

    This paper is organized as follows. In Section 2, we review the generalized fractional operators (integral and derivative operators). In Section 3, we transform TVPs for systems of FDEs into weakly singular Fredholm-Volterra integral equations with vanishing delay. In Section 4, we obtain a lower bound for the well-posedness of such problems. In Section 5, we propose a numerical method, and in Section 6 we provide an error analysis for the proposed method. In Section 7, after validating the proposed method, we compare the effects of various choices of weight function on modeling with a TVP.

    The generalized fractional operator is defined with respect to the weight function g:[a,b]R with the following properties:

    C1: gC1[a,b];

    C2: g is a strictly increasing function;

    C3: g1:g([a,b])[a,b] exists and is continuous (by C2, g([a,b])=[g(a),g(b)]).

    Definition 2.1. [18] Let fC[a,b] (a,bR). The generalized fractional integral GaIα,gt of order α (αC, Re(α)>0) with the weight function g is defined by

    GaIα,gtf(t)=1Γ(α)tag(τ)f(τ)(g(t)g(τ))1αdτ, t[a,b]. (2.1)

    Remark 1. Particular choices of g result famous generalizations and classical definitions of fractional operators:

    (I). If g(x)=x, then GaIα,gt is the classical Riemann-Liouville fractional integral,

    (II). If g(x)=ln(x) on [1,b], (b>1), then GaIα,gt is the Hadamard fractional integral, in this case g1(x)=ex,

    (III). If g(x)=eλxλ (λ>0) on [a,b], then GaIα,gt is the exponential fractional integral, in this case g1(x)=ln(λx)λ, [16],

    (IV). If g(x)=xρ, then GaIα,gt is the Katugampola fractional integral, in this case g1(x)=x1ρ [14,15].

    Definition 2.2. Let X[a,b] be a space of real valued functions on [a,b]. The weighted space with respect to the weight function g is defined by

    Xg[a,b]={k:[a,b]Rkog1X[g(a),g(b)]},

    with the weighted norm

    kg=k(g1),

    where . is a norm of the space X[g(a),g(b)].

    For example, Lg[a,b] and Cg[a,b] are weighted spaces of the Lebesgue integrable functions and the continuous functions with Lebesgue norm and supremum norm, respectively.

    Remark 2. Suppose kC[a,b]. Since g1 is continuous, kog1 is continuous. Thus, kCg[a,b] and C[a,b]Cg[a,b]. Similarly, we can prove that Cg[a,b]C[a,b]. For supremum norm in this space, we have

    sup[a,b]|k(x)|=sup[g(a),g(b)]|k(g1(x))|.

    Thus, Cg[a,b]=C[a,b]. However, this type of equivalency can not happen for all spaces like L[a,b]. In this case, by substituting x=g(y) and dx=g(y)dy, we have

    kg=k(g1)=g(b)g(a)|k(g1(x))|dx=ba|k(y)|g(y)dyk.

    Throughout the paper we generally use the supremum norm unless we mention it. The following theorem describes the generalized fractional integral by the Riemann-Liouville fractional integral.

    Theorem 2.1. Suppose αC, Re(α)>0, and g:[a,b]R has the properties C1C3. Then, for hC[a,b],

    GaIα,gth(g(t))=(RLg(a)Iαth(t))o(g(t)), (2.2)

    where RLg(a)Iαt stands for the Riemann-Liouville fractional integral, and the notation "o" stands for the combination operator.

    Proof. Setting k(t)=h(g(t)), we get

    GaIα,gtk(t)=1Γ(α)tah(g(τ))g(τ)(g(t)g(τ))1αdτ=1Γ(α)g(t)g(a)h(x)(g(t)x)1αdx =(1Γ(α)tg(a)h(x)(tx)1αdx)o(g(t)). (2.3)

    Here, we integrated with substitution x=g(τ), (dx=g(τ)dτ). Thus, we obtain

    GaIα,gtk(t)=(RLg(a)Iαth(t))o(g(t)).

    This completes the proof.

    Remark 3. Theorem 2.1 shows that RLg(a)Iαtk(g1(t)) is well-defined if and only if GaIα,gtk(t) is well-defined. Let X[g(a),g(b)] be the space that the operator RLg(a)Iαt is well-defined. Then, the operator GaIα,gtk(t) is well-defined on the weighted space Xg[a,b].

    Definition 2.3. [18] Let αC, Re(α)>0, and n=[Re(α)]+1. The generalized fractional derivatives on [a,b] (0a<b) are defined by

    GaDα,gtf(t)=(GaInα,gt(1g(t)ddt)nf(t))=1Γ(nα)ta(1g(τ)ddτ)nf(τ)(g(t)g(τ))1n+αg(τ)dτ (2.4)

    for αN and for a Borel function f:[a,b]R in which the integral in Eq (2.4) is well-defined. For αN, it is the classical integer order definition, i.e., KaDα,ρtf(t)=dn1dtn1.

    Theorem 2.2. Let αC, and 0<Re(α)<1. Then,

    GaDα,gtf(g(t))=(Cg(a)Dαtf(t))o(g(t)). (2.5)

    Proof. Setting h(t)=f(g(t)), we get

    GaDα,gth(t)=1Γ(1α)tah(τ)(g(t)g(τ))αdτ=1Γ(1α)taf(g(τ))g(τ)(g(t)g(τ))αdτ=1Γ(1α)g(t)g(a)f(y)(g(t)y)αdy=(Cg(a)Dαtf(t))o(g(t)). (2.6)

    This completes the proof.

    Computations of generalized derivatives and integrals of functions (g(t)g(a))v, for 0<Re(α)<1 are a direct consequence of Theorems 2.1 and 2.2. Thus, we have

    GaDα,gt(g(t)g(a))v=(Cg(a)Dαt(tg(a))v)(g(t))={(Γ(v+1)Γ(vα+1)(tg(a))vα)o(g(t)),v0,0,v=0,={(Γ(v+1)Γ(vα+1)(g(t)g(a))vα),v0,0,v=0, (2.7)

    and

    GaIα,gt(g(t)g(a))v=(RLg(a)Iαt(tg(a))v)o(g(t))=(Γ(v+1)Γ(v+α+1)(tg(a))v+α)o(g(t))=Γ(v+1)Γ(v+α+1)(g(t)g(a))v+α. (2.8)

    Now, we can use Theorem 2.2 to find a suitable space for defining a generalized fractional derivative. Regularly, the spaces ACn[a,b] and Cn[a,b] are the best choices for classical fractional derivatives with order α such that n1<Re(α)<n, (Theorems 2.1 and 2.2 of [28], also see [13]). We recall that

    ACn[a,b]={f:[a,b]C,f(n1)AC[a,b]},  nN,

    where AC[a,b] is the set of absolute continuous functions. Thus, we can state the following lemma.

    Lemma 2.1. Let fACng[a,b]. Then, GaDα,gtf(t) exists almost everywhere on [a, b]. Moreover, if fCng[a,b], then GaDα,gtfC[a,b].

    Proof. The proof is a straightforward result of Eq (2.5), and Theorems 2.1 and 2.2 of [28].

    Remark 4. Equations (2.2) and (2.5) can be rewritten as

    GaIα,gtf(t)=(RLg(a)Iαtf(g1(t)))o(g(t)) (2.9)

    and

    GaDα,gtf(t)=(Cg(a)Dαtf(g1(t)))o(g(t)). (2.10)

    Lemma 2.2. Let α(0,1], ρ[0,1), and fAC[a,b].Then,

    GaItα,gGaDα,gtf(t)=f(t)f(a). (2.11)

    Proof. It follows from Eq (2.9) that

    GaItα,g(GaDα,gtf(t))=(RLg(a)Iαt(GaDα,gtf(t)o(g1(t))))o(g(t)).

    Applying Eq (2.10), we get

    GaItα,g(GaDα,gtf(t))=(RLg(a)Iαt((Cg(a)Dαtf(g1(t)))o(g(t))o(g1(t))))o(g(t)).

    Taking into account that (g(t))o(g1(t))=t, we obtain

    GaItα,g(GaDα,gtf(t))=(RLg(a)Iαt(Cg(a)Dαtf(g1(t))))o(g(t)).

    Finally, from properties of classical fractional operators, we obtain

    GaItα,g(GaDα,gtf(t))=(f(g1(t))f(g1(g(a))))o(g(t))=(f(g1(t))f(a))o(g(t))=f(t)f(a), (2.12)

    which completes the proof.

    Remark 5. All the mentioned results can be generalized to higher-dimensional spaces by the vector order notation. For vector order ˉα=[α1,,αν]T (Re(αi)(0,1), i=1,,ν) and vector weights g=[g1,,gν]T we have

    GaItˉα,gGaDˉα,gtf(t)=[GaItα1,g1GaDα1,g1tf1,,GaIαν,gνGaDαν,gνtfν]T=[f1(t)f1(a),,fν(t)fν(a)]T=[f1(t),,fν(t)]T[f1(a),,fν(a)]T=f(t)f(a), (2.13)

    where f=[f1,,fν]T belongs to the ν-dimensional space (AC[a,b])ν. It is a straightforward generalization of a one-dimensional case since all operators operate element-wise [21].

    As an application of Eqs (2.7), (2.8)and (2.11), we can solve a linear initial value problem for FDEs:

    GaDˉα,gty=Ay(t)+B, t(a,b), (2.14)

    with the initial value y(0)=y0 and imposing αi=α, gi=g for i=1,,ν. Applying GaItα,g to both sides of Eq (2.14) and using Eq (2.11), we have

    y=y0+GaItˉα,gB+GaItˉα,gAy(t), t(a,b). (2.15)

    Here, it is worth to mentioning that, for general ˉα and g, the commutative property

    GaItˉα,gAyn=GaItˉα,gAyn

    does not hold for vector order fractional integrals as well as for vector order fractional derivatives.

    The iterative method

    yn+1=y0+GaItˉα,gB+GaItˉα,gAyn,  n=1,2,..., (2.16)

    with the initial value

    y1=y0+GaItˉα,gB

    is utilized to obtain the exact solution y=limnyn. Recursively, we obtain

    y2=y0+GaItˉα,gB+AGaItˉα,gy1=y0+BΓ(ˉα+1).(g(t)g(a))ˉα+AGaItˉα,g(y0+GaItˉˉα,gB)=y0+BΓ(ˉα+1).(g(t)g(a))ˉα+Ay0Γ(ˉα+1).(g(t)g(a))ˉα+ABΓ(2ˉα+1).(g(t)g(a))2ˉα (2.17)

    and similarly

    y3=y0+GaItˉα,gB+AGaItˉα,gy2=y0+BΓ(ˉα+1).(g(t)g(a))ˉα+Ay0Γ(ˉα+1).(g(t)g(a))ˉα+ABΓ(2ˉα+1)(g(t)g(a))2ˉα+A2y0Γ(2ˉα+1).(g(t)g(a))2ˉα+A2BΓ(3ˉα+1).(g(t)g(a))3ˉα. (2.18)

    This pattern suggests

    y(t)=n=0Any0Γ(nˉα+1).(g(t)g(a))nˉα+AnBΓ((n+1)ˉα+1).(g(t)g(a))(n+1)ˉα (2.19)

    where the dot (".'') operator stands for element-wise multiplication.

    Applying the fractional integral to both sides of the system (1.2), it follows from Eq (2.2) that

    y(t)=y(a)+GaItˉα,gf(t,y(t)), t(a,b), (3.1)

    or equivalently,

    y(t)=y(a)+RLg(a)Iˉαtf(g1(t),y(g1(t)))og(t). (3.2)

    Computationally, it is important to note that

    f(g1(t),y(g1(t)))=[f1(g11(t),[y1(g11(t)),,yν(g11(t))])fν(g1ν(t),[y1(g1ν(t)),,yν(g1ν(t))])].

    Putting t=b in Eq (3.2) and using the terminal condition, we obtain

    y(a)=y(b)RLg(a)Iˉαtf(g1(t),y(g1(t)))og(t)t=b. (3.3)

    Substituting y(a) from Eq (3.3) into Eq (3.2), we obtain

    y(t)=y(b)RLg(a)Iˉαtf(g1(t),y(g1(t)))og(t)t=b+RLg(a)Iˉαtf(g1(t),y(g1(t)))og(t). (3.4)

    Equation (3.4) can be represented by a weakly singular Fredholm-Volterra integral equation with vanishing delay g(t)

    y(t)=yb1Γ(ˉα)g(b)g(a)f(g1(τ),y(g1(τ)))(g(b)τ)1ˉαdτ+1Γ(ˉα)g(t)g(a)f(g1(τ),y(g1(τ)))(g(t)τ)1ˉαdτ, (3.5)

    or equivalently,

    y(t)=yb1Γ(ˉα)bag(x)f(x,y(x))(g(b)g(x))1ˉαdx+1Γ(ˉα)tag(x)f(x,y(x))(g(t)g(x))1ˉαdx. (3.6)

    Remark 6. If g(t)=pt, the corresponding system has proportional delay [29]. Also, with g(t)=t32 and the non-local boundary condition

    y(b)=1+1Γ(ˉα)bag(x)f(x,y(x))(g(b)g(x))1ˉαdx,

    the system (3.6) is a Lighthill system. This system is used for describing the temperature distribution of the surface of a projectile moving through a laminar layer [30].

    We suppose the vector function f=[f1,,fν]T satisfies the following conditions.

    (H1) The vector function f is continuous with respect to its variables on [0,T]×Rν,

    (H2) Functions fi, i=1,,ν, are Lipschitz functions with Lipschitz constants Li, i.e.,

    fi(t,w)fi(t,y)Liwy

    for y,wRν, and we set

    LM=maxi=1,,νLi.

    It is a straightforward conclusion of Theorem 2.1 and Remark 2 that the operator

    GaItˉα,gf(t,.):(C[a,b])ν(C[a,b])ν

    transforms a continuous vector function into a continuous vector function, and thus it is well-defined.

    We use the norm

    y=maxi=1,,νyi,yi=supt[a,b]y(t)

    to state continuity of the operator GaItˉα,gf(t,.).

    Theorem 4.1. Let assumptions (H1) and (H2) hold for the vector function f. Then, the operators GaItˉα,gf(t,.) and GaIbˉα,gf(b,.) are continuous, i.e.,

    GaItˉα,gf(t,u)GaItˉα,gf(t,v)Muv (4.1)

    and

    GaIbˉα,gf(b,u)GaIbˉα,gf(b,v)Muv, (4.2)

    where the vector functions u and v belong to (C[a,b])ν and

    M=maxi=1,,νLi(gi(b)gi(a))ˉαΓ(ˉα+1). (4.3)

    Proof. The supremum norm of the ith component of the generalized fractional integral operator can be estimated by

    (GaItˉα,gf(t,u))i(GaItˉα,gf(t,v))i=supt[a,b]1Γ(αi)ta|gi(x)(fi(x,u(x)fi(x,v(x))|(gi(t)gi(x))1αidx.

    By assumptions C1C3, gi(x) and (gi(t)gi(x))1αi for x[0,t] are non-negative. Thus,

    (GaItˉα,gf(t,u))i(GaItˉα,gf(t,v))i=supt[a,b]LiΓ(αi)tagi(x)(gi(t)gi(x))1αidxuv=LiΓ(αi)supt[a,b](gi(t)gi(a))αiαiuv=Li(gi(b)gi(a))αiΓ(αi+1)uv. (4.4)

    Consequently, the inequalities (4.1) and (4.2) follow from Eq (4.4).

    By Theorem 4.1, the operator T:(C[a,b])ν(C[a,b])ν described by

    T(y)=ybGaIbˉα,gf(b,y)+GaItˉα,gf(t,y) (4.5)

    is well-defined.

    Theorem 4.2. Let assumptions (H1) and (H2) hold for f:[0,T]×RνRν.Let

    M:=maxi=1,,ν2Li(gi(b)gi(a))αiΓ(αi+1)<1. (4.6)

    Then, the system (3.6) has a unique solution on (C[a,b])ν.

    Proof. Systems (3.5) and (3.6) can be rewritten in the operator form

    y=T(y). (4.7)

    By using Theorem 4.1 for the vector functions u and v(C[a,b])ν, we have

    T(u)T(v)Muv. (4.8)

    From condition (4.6), M<1, and hence

    T(u)T(v)uv. (4.9)

    Thus, T is a contracting operator, and by the Banach fixed point theorem, it has a unique solution on (C[a,b])ν. This completes the proof.

    Let us adopt notations of [29]. Let Ih={tn: a=t0<t1<<tN=b} be a partition of [a,b], hn=tn+1tn, σn=(tn,tn+1] (n=0,,N1) and h=maxn=0,,N1hn. We recall that the grading mesh points of the form tn:=(ba)(n/N)r are an appropriate choice for integral equations with weakly singular kernels[29]. The piecewise polynomial space can be defined by

    S1m1(Ih)={v:vσnΠm1},

    where Πm1 is the space of polynomials of degree less than m. An approximate solution of the system (1.2) has the dense representation

    uN(t0)=y0,uN(tn+vhn)=mi=1Li(v)Un,i,v(0,1], n=0,,N1, (5.1)

    where Li are Lagrange interpolation formulas, and Un,i:=uN(tn,i) and collocation points are defined by tn,i:=tn+cihn for given collocation parameters 0c1<<cm1. The operator PN that projects (C[0,T])ν into (S1m1(Ih))ν is described by

    PN(u(tn,i))=u(tn,i), i=1,,m, n=0,,N1.

    Thus, we can explicitly and uniquely define PN by

    PN(uN(tn+vhn))=mi=1Li(v)Un,i, n=0,,N1. (5.2)

    The quadrature approximation of the operator GaItˉα,gf(t,.) is computed by

    QNuN(t)=GaItˉα,gPNf(t,uN(t)).

    Let t:=tn+vhn[tn,tn+1]. Then,

    QNuN(t)=n1l=0mj=1λn,l,j(v)f(tl,j,Ul,j)+mj=1λn,n,j(v)f(tn,j,Un,j), (5.3)

    where the weights of the quadrature are determined by

    λn,l,j(v)={hlΓ(ˉα)10g(tl+zhl)Lj(z)(g(tn+vhn)g(tl+zhl))1ˉαdz,l=0,,n1,hnΓ(ˉα)v0g(tn+zhn)Lj(z)(g(tn+vhn)g(tn+zhn))1ˉαdz,l=n, (5.4)

    for j=1,,m and n=0,,N1. Similarly, a quadrature method for GaIbˉα,gf(t,.) can be computed by

    ˜QNuN(t)=GaIbˉα,gPNf(t,uN(t))

    or, equivalently,

    ˜QNuN=N1l=0mj=1λN1,l,j(1)f(tl,j,Ul,j). (5.5)

    Setting t=to,k for o=0,,N1 and k=1,,m, and using the introduced quadrature method, the unknown Uo,k can be obtained by solving the system of nonlinear equations described by

    Uo,k=yb˜QNuN+QNuN(to,k). (5.6)

    After solving this system, we can obtain the dense approximate solution by Eq (5.1). The nonlinear system (5.6) is solved by the recursive formula

    Uio,k=yb˜QNui1N+QNUi1o,k,  i=1,...,nt, (5.7)

    where nt is the constant number of iterations (can be chosen by a user), or can be adopted by

    uiNui1NTol

    with a given tolerance Tol.

    In this research, we also need to introduce a numerical method for solving related initial value problems. Let us consider the system (1.2) with a given initial condition

    y(a)=ya. (5.8)

    A numerical approach by recursively solving the νm dimensional algebraic systems

    Uo,k=ya+QNuN(to,k), k=1,,m, (5.9)

    for o=0,,N1 is utilized to obtain the corresponding dense solution. However, for the IVP in each interval, we need to solve only a nonlinear equation of dimension ν×m, while for the TVP we need to solve a nonlinear equation of dimension ν×m×N.

    Does the nonlinear system (5.6) have a solution? Is it unique? What is the convergence of the proposed iterative method (5.7)? This subsection answers these questions.

    Theorem 5.1. Let assumptions (H1) and (H2) hold for the vector function f:[0,T]×RνRν. Let

    Λ:=maxi=1,,ν2Li(gi(b)gi(a))αiPN2Γ(αi+1)<1 (5.10)

    for a given N. Then, the approximate solution of the system (5.6) exists and is unique, and the iterative method (5.7) converges to the solution of the system (5.6) on collocation points.

    Proof. Since uN(S1m1(Ih))ν we have PNuN=uN, and the dense approximate solution (5.1) satisfies

    uN(t)=PNTNuN, (5.11)

    where

    TNu:=yb˜QNu+QNu.

    Let u,v(S1m1(Ih))ν. The ith component of the QNu(t)QNv(t) satisfies

    |(QNu(t))i(QNv(t))i|=|(GaItˉα,gPNf(t,u(t)))i(GaItˉα,gPNf(t,v(t)))i||1Γ(αi)ta|gi(x)(PN(f(x,u(x)f(x,v(x)))i|(gi(t)gi(x))1αidx||1Γ(αi)tagi(x)dx(gi(t)gi(x))1αi|(PN(f(x,uf(x,v))i(gi(t)gi(a))αiΓ(αi+1)Li(PN)i(uv)iLi(gi(t)gi(a))αiΓ(αi+1)PNuv. (5.12)

    Therefore,

    QNuQNvMPNuv, (5.13)

    where M is defined as in Eq (4.3). Similarly,

    ˜QNu˜QNvMPNuv. (5.14)

    The operator PNTN is a contractor since

    PNTN2PN2M(uv)uv

    by the hypothesis of the theorem. It follows from the Banach fixed point theorem that the operator

    PNTN:(S1m1(Ih))ν(S1m1(Ih))ν

    has a unique solution, and the recursive formula

    uiN=PNTNui1N, i=1,2,.

    converges to the fixed point of PNTN with any initial solution

    u0N(S1m1(Ih))ν.

    We recall that (S1m1(Ih))ν is a finite dimensional space and thus a Banach space.

    In this section, we show that the error of the discretized collocation method is proportional to the error of the quadrature method for the exact solution:

    R(y):=(TPNTN)(y). (6.1)

    Theorem 6.1. Let assumptions (H1) and (H2) hold for f:[0,T]×RνRν. Let Λ defined by (5.10) satisfy

    Λ<1. (6.2)

    Then, the approximate solution described by (5.1) and (5.6) exists, and

    uNy11CR(y). (6.3)

    Proof. Since uN(S1m1(Ih))ν, it follows that PNuN=uN. Hence, the dense approximate solution (5.1) satisfies

    uN(t)=PNTNuN, (6.4)

    where

    TNu=yb˜QNu+QNu(t).

    By using Eq (6.1),

    y=PNTN(y)+R(y). (6.5)

    Setting

    e:=uNy

    and subtracting Eq (6.5) from Eq (5.10), we obtain

    e=PNTNuNPNTN(y)R(y). (6.6)

    The first difference of Eq (6.6) satisfies

    PNTNuNPNTN(y)PNTNuNTN(y)PN˜QNuN+QNuN(t)+˜QNyQNy(t)PN(˜QNuN˜QNy+QNuN(t)QNy(t)). (6.7)

    Let us consider the the ith component of the QNuN(t)QNy(t), and we have

    |(QNuN(t))i(QNy(t))i|=|(GaItˉα,gPNf(t,uN(t)))i(GaItˉα,gPNf(t,y(t)))i||1Γ(αi)ta|gi(x)(PN(f(x,uN(x)f(x,y(x)))i|(gi(t)gi(x))1αidx||1Γ(αi)tagi(x)dx(gi(t)gi(x))1αi|(PN(f(x,uN(x)f(x,y(x)))i(gi(t)gi(a))αiΓ(αi+1)Li(PN)i(uNy)iLi(gi(t)gi(a))αiΓ(αi+1)PNuNy. (6.8)

    It follows from Eq (6.8) that

    QNuNQNyMPNuNy. (6.9)

    Similarly, we can prove that

    ˜QNuN˜QNyMPNuN(x)y(x). (6.10)

    Substituting Eqs (6.9) and (6.10) into Eq (6.7), we obtain

    PNTNuNPNTN(y)2MPN2uN(x)y(x)=2MPN2e. (6.11)

    Taking the norm from Eq (6.6) and using Eq (6.11), we obtain

    ePNTNuNPNTN(y)+R(y)2MPN2e+R(y). (6.12)

    This leads to an interesting bound for the error,

    e11ΛR(y). (6.13)

    The following lemma supports Theorem 6.1 by a bound for PN.

    Lemma 6.1. Let PN:(C[0,T])ν(S1m1(Ih))ν be a projection defined by Eq (5.2). Let . be the induced norm of projection. Then, there exist N0N and a constant C such that

    PNC

    for all N>N0.

    Proof. We prove this statement for the 1-D case, and the proof of the ν-dimensional case is similar. Let y<1.

    PNy=maxn=0,,N1PNyσn={maxt[tn,tn+1], n=0,,N1mj=1mi=1,ij|y(tn,j)|,m=1,maxt[tn,tn+1], n=0,,N1mj=1mi=1,ijttn,itn,jtn,i|y(tn,j)|,m>1.

    Setting t=tn+vhn and |y(z)|=max|y(tn,j)|, we obtain

    PNyC|y(z)|Cy

    where

    C={1,m=1,maxv[0,1]mj=1mi=1,ij|vci||cjci|,m>1,

    is independent of y and t.

    Remark 7. We note that for m=1, we have PN1. Thus, Λ in Eqs (6.2) and (5.10) will be equal to M defined in (4.6). This means that the conditions of (6.1), (5.1) and (4.2) coincide.

    To start numerical implementation, we compute the weight coefficients for the given parameters m and c. Let m=1 and c1=θ, L1(t)=1. The corresponding methods are known as the θ method. The θ method approximates the solution with a constant function in each interval. In uniform mesh, it is similar to Haar scale functions. The weight of the proposed approximate quadrature can be computed by

    λn,l,1(v)={(g(tn+vhn)g(tl))ˉαΓ(ˉα+1)(g(tn+vhn)g(tl+1))ˉαΓ(ˉα+1),l=0,,n1,(g(tn+vhn)g(tl))ˉαΓ(ˉα+1),l=n.

    Consider a system of GFDEs described by Eq (1.2) with the vector order ˉα=[0.5,0.5]T, the weight function g=[g,g]T, the source function

    f(t,y)=[0.1sin(y21+y22)0.1sin(2(g(t)g(a)))+π20.1cos(y1y2)+0.1π20.1]

    and the terminal value

    yb=[(g(b)g(a)),(g(b)g(a))]T.

    Before giving the exact solution, we notice that f satisfies the condition (H1) and (H2). According to Theorem 4.2, for an appropriately chosen value of b, the solution falls into (C[a,b])2. The exact solution of this system is

    y=[(g(t)g(a)),(g(t)g(a))]T.

    This is in agreement with the assertion of Theorem 4.2.

    We notice that we could not choose a=0 for case II, since the function ln is not defined at zero. For a better comparison, we choose the initial terminal a=1, and we find b such that

    bg1(π0.16). (7.1)

    Equation (7.1) guarantees the condition (4.6) of Theorem 4.1. Thus, the corresponding system has a unique solution with an arbitrary value of yb. Table 1 shows lower bounds of b for various choices of weight functions g obtained by Eq (7.1).

    Table 1.  The bounds for different choices of weight function g.
    Category type Value of parameters The bound
    I no parameter 20.6350
    II no parameter 3.3678e+08
    III λ=0.1 11.2124
    III λ=1 3.1070
    III λ=2 1.9214
    IV ρ=0.1 1.3997e+13
    IV ρ=0.5 425.8013
    IV ρ=10 1.3535

     | Show Table
    DownLoad: CSV

    Remark 8. Table 1 highlights case II (Hadamard fractional operator) as well as case IV (Katugampola type) as candidates for having bigger bounds of well-posedness. The parameters of cases III and IV change b with different rates. The most rapid change occurs in the Katugampola type. We recall that the terminal value problem may not be well-posed [4], at all.

    Remark 8 guides us toward available options for modeling inverse problems in applied science. Actually, we need well-posedness for larger intervals in general.

    In Figure 1, we illustrate the exact and the approximate solutions for various weight functions g, with the upper bound b=2, the collocation parameter c=0.1, the graded mesh exponent r=2 and the number of steps N=20. The maximum errors and estimated orders of convergence for both components of the solution are compared in Tables 2 and 3, respectively. The maximum error is computed by

    EN=maxn=0,,N1{uN(tn,1)y(tn,1)}.
    Figure 1.  A comparison of the exact and the approximate solutions for various values of weight functions g.
    Table 2.  Maximum error and the estimated order of convergence for the first component of the solution.
    Category type E32 E64 E128 log2E32/E64 log2E64/E128
    I 0.0313 0.0156 0.0079 1.0000 0.9773
    II 0.0312 0.0156 0.0079 0.9997 0.9772
    III, λ=0.1 0.0353 0.0177 0.0090 0.9981 0.9771
    IV, ρ=0.1 0.0099 0.0049 0.0025 0.9998 0.9772

     | Show Table
    DownLoad: CSV
    Table 3.  Maximum error and the estimated order of convergence for the second component of the solution.
    Category type E32 E64 E128 log2E32/E64 log2E64/E128
    I 0.0313 0.0156 0.0079 1.0000 0.9773
    II 0.0312 0.0156 0.0079 0.9997 0.9772
    III, λ=0.1 0.0353 0.0177 0.0090 0.9981 0.9771
    IV, ρ=0.1 0.0099 0.0049 0.0025 0.9998 0.9772

     | Show Table
    DownLoad: CSV

    It is known that, for y(C[a,b])2, R(y)=O(Nm)+O(Nrβ) (see for example [9]). In this example, m=1 and r=2; thus, according to Theorem 6.1,

    e=O(N1).

    Tables 2 and 3 show that the order of converge tends to 1. Thus, the reported results are in complete agreement with the theoretical results.

    Diabetes is a silent epidemic of high blood sugar levels, mainly due to decreasing activities and an inappropriate diet. An epidemiology model using FADEs for simulating the population of diabetes cases is proposed in [31]. This model applies linear two-dimensional FDEs

    GaD[α1,α2]T,[g1,g2]Tty(t)=Aλ(t)+B (7.2)

    to describe the population of diabetes cases. Here,

    A=[a11a12a21a22]

    is a 2×2 matrix, and B is a 2-dimensional column vector. By some physical interpretation of the model [31], we impose 0a11<5, 0a121 and 0a21<2. The interpretation of a22 is the rate of death in [31]. However, the population of the studied zone is not a constant parameter, and indeed it is increasing. Thus, we let a22 be positive, and we impose 1a22<8.

    Let us use the system (7.2) as a black box for modeling the populations of diabetes cases with model parameters A and B. We use available data, which was reported in [32], for the population of diabetes cases from 19902017 to find the model parameters. Then, we use the obtained model for predicting the past dynamics by a terminal value problem.

    We set ˉα=[0.9,0.9]T to involve history in our model. Our interest is to study the effect of g on the results of this model. Let the output of the black box be o(t)=y2(t), indicating the population of diabetes cases for an input value of t.

    We use o for simulating the dynamics of the prevalence number of diabetes. Let D(t) show available data for the prevalence number of diabetes. Then, D(t)/DM is normalized data, where DM=max|D(t)|.

    Setting A=[0,0;0,0], we obtain

    y(t)=y0+(g(t)g(a))αΓ(α+1)B.

    This means, by tending α0, the matrix B can be regarded as an approximation for the slope of the curve y(t). Thus, a reasonable physical interpretation suggests choosing B as the prevalence rate of diabetes with some gain related to g.

    According to [32], a prevalence rate is a number between 4500 and 5500 per 100k population. Thus, we use B=η[0,0.045]T to reduce the parameters in the related optimization problem, where the parameter ηR is a model parameter.

    Now, the inverse problem of modeling dictates to determining A and η such that minimizes the following objective function:

    E=2017t=1990o(t)D(t)/DM. (7.3)

    The output o(t) will be calculated from the block box as an initial value problem on [1990,2017]. The modeled IVP can be solved by Eq (2.19) or a numerical method described by (5.9). We applied the latter one.

    Let's apply the inverse problem to the obtained model. We emphasize that classical fractional-order TVPs (i.e., TVPs with Caputo fractional derivative) for high dimensional systems are not well-posed on the large intervals. A cure for such a modeling problem is the ultimate aim of this paper. To this end, we use appropriate generalized fractional derivatives with a larger well-posedness bound.

    Let g(x)=[xρt,xρt]T. The case ρ=1 corresponds to the Caputo fractional derivative. In Figure 2(a) we depict the result of the fitted model for ρ=1. We apply (5.9) with N=200, c=1 and graded mesh of exponent r=2. The fitting error (7.3) in this case is E=0.24998. Since Λ=113.2473, the conditions (6.2) and (4.6) do not hold. Thus, Theorems 4.2, 5.1 and 6.1 do not guarantee the corresponding numerical and exact solutions exist. Also, there is no guarantee for convergence of the approximate solution. Figure 2(b) shows the approximate solution by the prescribed method in Eqs (5.1) and (5.7) (N=200, c=1 and r=2) for the corresponding TVP. It is not a surprise that the numerical approximation for the TVP does not converge to the corresponding solution of the IVP. Thus, this model cannot be used for predicting past dynamics.

    Figure 2.  FDEs (7.2) with g(x)=[x,x]T and (a) normalized data and simulated results with the initial condition, (b) normalized data and simulated results with terminal conditions. The solution of the TVP does not converge with the solution of the IVP.

    Mathematically, the constraints for elements of A impose that the Lipschitz constant of f be less than 8. As a result, we can improve the domain of well-posedness by decreasing ρ in g(x)=[xρt,xρt]T. On the other hand, this restriction may increase the optimization error, i.e., E. Using the same procedure as we reported for the case ρ=1, we report the result for ρ=0.2,0.4,0.8 in Figures 35, respectively (the case ρ=0.6 is similar to case ρ=0.8 and it does not add more information).

    Figure 3.  FDEs (7.2) with g(x)=[xρ,xρ]T and ρ=0.8. (a) Modeling, (b) Past dynamics (validation) and (c) Past dynamics (prediction).
    Figure 4.  FDEs (7.2) with g(x)=[xρ,xρ]T and ρ=0.4. (a) Modeling, (b) Past dynamics (validation) and (c) Past dynamics (prediction).
    Figure 5.  FDEs (7.2) with g(x)=[xρ,xρ]T and ρ=0.2. (a) Modeling, (b) Past dynamics (validation) and (c) Past dynamics (prediction).

    For the case ρ=0.8, the fitting error E=0.1530 is decreased (Figure 3(a)(c)). The λ corresponding to interval I1:=[1990,2017] is decreased to 13.1412. It is not in the zone of the predicted bound for well-posedness and convergence of the given TVP. However, it converges, and the solutions of the TVP and IVP coincide, as shown in Figure 3(b). The solution for prediction of past time on I2=[1940,2017] with the same TVP is illustrated in Figure 3(c). Obviously, increasing the interval length increases λ. For interval I2 we get λ=33.8248. There is no guarantee of well-posedness, and the results cannot be reliable (a similar pattern repeats for the case λ=0.6).

    For the case ρ=0.4, the fitting error E=0.37360 is increased (Figure 4(a)(c)). However, the values λ corresponding to I1 and I2 are decreased, and both are near the well-posedness and convergence zone. We can rely on these results.

    The last case is ρ=0.2: see Figure 5(a)(c). The optimization result will cross some boundaries of our restricted conditions on A. The values of λ for both intervals are less than 1. Well-posedness of the corresponding TVP and convergence of the proposed numerical method are ensured. However, the error decreases to E=0.9076. It behaves not better than a model with linear regression.

    Conclusively, there is a contrast between the fitting error and the well-posedness condition. A compromise can be made by considering both conditions for a given model. In this case, we propose the model with ρ=0.4, for estimating past dynamics.

    An extensive application of FDEs is involved in many branches of science, such as biology, epidemiology, medicine and chemistry. Always, these systems are higher dimensional and generally involve some terms of nonlinear Lotka Volterra dynamics with initial conditions. Therefore, as IVPs they are well-posed. However, the inverse problem is not well-posed for higher dimensional equations. It is useless and dangerous to use fractional derivatives for modeling such processes, especially when we have boundary conditions, without considering their well-posedness. Fortunately, this paper's results for modeling with general fractional derivatives open a new window to overcome this problem. However, we encounter another problem: increasing fitting error. For this aim, a compromising solution between well-posedness and less fitting error is considered.

    The authors declare no conflict of interest.



    [1] R. P. Agarwal, M. Bohner, S. R. Grace, et al. Discrete Oscillation Theory, Hindawi Publishing Corporation, New York, 2005.
    [2] M. Bohner, T. Li, Oscillation of second order p-Laplace dynamic equations with a nonpositive neutral coefficient, Appl. Math. Lett., 37 (2014), 72-76. doi: 10.1016/j.aml.2014.05.012
    [3] Y. Huang, J. Wang, T. Li, Oscillation of second order difference equations, J. Nonlinear Sci. Appl., 10 (2017), 1238-1243. doi: 10.22436/jnsa.010.03.32
    [4] V. Lakshmikantham, D. D. Bainov, P. S. Simieonov, Oscillation Theory of Impulsive Differential Equations, World Scientific, Singapore, 1989.
    [5] Q. L. Li, Z. G. Zhang, F. Gou, et al. Oscillation criteria for third-order difference equations with impulses, J. Comput. Appl. Math., 225 (2009), 80-86. doi: 10.1016/j.cam.2008.07.002
    [6] T. Li, S. H. Saker, A note on oscillation criteria for second-order neutral dynamic equations on isolated time scales, Commun. Nonlinear Sci. Numer. Simul., 19 (2014), 4185-4188. doi: 10.1016/j.cnsns.2014.04.015
    [7] W. Lu, W. G. Ge, Z. H. Zhao, Oscillatory criteria for third-order nonlinear difference equations with impulses, J. Comput. Appl. Math., 234 (2010), 3366-3372. doi: 10.1016/j.cam.2010.04.037
    [8] M. Peng, Oscllation theorems for second order nolnlinear neutral dealy difference equations with impulses, Comput. Math. Appl., 44 (2002), 741-749. doi: 10.1016/S0898-1221(02)00187-6
    [9] M. Peng, Oscillation criteria for second-order impulsive delay difference equations, Appl. Math. Comput., 146 (2003), 227-235.
    [10] I. Stamova, G. Stamov, Applied Impulsive Mathematical Models, MS Books in Mathematics, Springer, Switzerland, 2016.
    [11] Y. Tian, Y. Cai, T. Li, Existence of nonoscillatory solutions to second-order nonlinear neutral difference equations, J. Nonlinear Sci. Appl., 8 (2015), 884-892. doi: 10.22436/jnsa.008.05.38
    [12] A. K. Tripathy, On the oscillation of second order nonlinear neutral delay difference equations, Electron J. Qual. Theory Differ. Equ., 11 (2008), 1-12.
    [13] A. K. Tripathy, S. Panigrahi, On the oscillatory behaviour of a class of nonlinear delay difference equations of second order, Indian J. Pure Appl. Math., 42 (2011), 27-40. doi: 10.1007/s13226-011-0002-z
    [14] A. K. Tripathy, Oscillation criteria for a class of first order neutral impulsive differential-difference equations, J. Appl. Anal. Comput., 4 (2014), 89-101.
    [15] A. K. Tripathy, G. N. Chhatria, Oscillation criteria for forced first order nonlinear neutral impulsive difference system, Tatra Mt. Math. Publ., 71 (2018), 175-193.
    [16] A. K. Tripathy, G. N. Chhatria, On oscillatory first order neutral impulsive difference equations, Math. Bohem., 2019, 1-15.
    [17] A. K. Tripathy, G. N. Chhatria, Oscillation criteria for first order neutral impulsive difference equations with constant coefficients, Differ. Equ. Dyn. Syst., 2019, 1-14.
    [18] G. P. Wei, The persistance of nonoscillatory solutions of difference equation under impulsive perturbations, Comput. Math. Appl., 50 (2005), 1579-1586. doi: 10.1016/j.camwa.2005.08.025
    [19] H. Zhang, L. Chen, Oscillations criteria for a class of second-order impulsive delay difference equations, Adv. Complex Syst., 9 (2006), 69-76. doi: 10.1142/S0219525906000677
    [20] C. Zhang, R. P. Agarwal, M. Bohner, et al. New oscillation results for second-order neutral delay dynamic equations, Adv. Difference Equ., 2012 (2012), 1-14. doi: 10.1186/1687-1847-2012-1
  • This article has been cited by:

    1. Guo-Cheng Wu, Babak Shiri, Qin Fan, Hua-Rong Feng, Terminal Value Problems of Non-homogeneous Fractional Linear Systems with General Memory Kernels, 2022, 1776-0852, 10.1007/s44198-022-00085-2
    2. Kamal Shah, Thabet Abdeljawad, Study of a mathematical model of COVID-19 outbreak using some advanced analysis, 2022, 1745-5030, 1, 10.1080/17455030.2022.2149890
    3. Li Tian, Ziqiang Wang, Junying Cao, A high-order numerical scheme for right Caputo fractional differential equations with uniform accuracy, 2022, 30, 2688-1594, 3825, 10.3934/era.2022195
    4. Mohamed Obeid, Mohamed A. Abd El Salam, Jihad A. Younis, Operational matrix-based technique treating mixed type fractional differential equations via shifted fifth-kind Chebyshev polynomials, 2023, 31, 2769-0911, 10.1080/27690911.2023.2187388
    5. Min Wang, Yeliz Karaca, Mati ur Rahman, Mi Zhou, A theoretical view of existence results by using fixed point theory for quasi-variational inequalities, 2023, 31, 2769-0911, 10.1080/27690911.2023.2167990
    6. F. Mohammadizadeh, S.G. Georgiev, G. Rozza, E. Tohidi, S. Shateyi, Numerical solution of ψ-Hilfer fractional Black–Scholes equations via space–time spectral collocation method, 2023, 71, 11100168, 131, 10.1016/j.aej.2023.03.007
    7. Tongke Wang, Sijing Liu, Zhiyue Zhang, Singular expansions and collocation methods for generalized Abel integral equations, 2023, 03770427, 115240, 10.1016/j.cam.2023.115240
    8. Hamid Reza Marzban, Atiyeh Nezami, A hybrid of the fractional Vieta–Lucas functions and its application in constrained fractional optimal control systems containing delay, 2024, 1077-5463, 10.1177/10775463241273027
    9. Hasib Khan, Jehad Alzabut, J.F. Gómez-Aguilar, Praveen Agarwal, Piecewise mABC fractional derivative with an application, 2023, 8, 2473-6988, 24345, 10.3934/math.20231241
    10. Wadhah Al‐sadi, Zhouchao Wei, Tariq Q. S. Abdullah, Abdulwasea Alkhazzan, J. F. Gómez‐Aguilar, Dynamical and numerical analysis of the hepatitis B virus treatment model through fractal–fractional derivative, 2024, 0170-4214, 10.1002/mma.10348
    11. Ruiqing Shi, Yihong Zhang, Stability analysis and Hopf bifurcation of a fractional order HIV model with saturated incidence rate and time delay, 2024, 108, 11100168, 70, 10.1016/j.aej.2024.07.059
    12. Babak Shiri, Yong-Guo Shi, Dumitru Baleanu, The Well-Posedness of Incommensurate FDEs in the Space of Continuous Functions, 2024, 16, 2073-8994, 1058, 10.3390/sym16081058
    13. A. M. Kawala, H. K. Abdelaziz, A hybrid technique based on Lucas polynomials for solving fractional diffusion partial differential equation, 2023, 9, 2296-9020, 1271, 10.1007/s41808-023-00246-4
    14. Hui Fu, Wei Xie, Yonggui Kao, Adaptive sliding mode control for uncertain general fractional chaotic systems based on general Lyapunov stability, 2024, 05779073, 10.1016/j.cjph.2024.05.032
    15. Bashir Ahmad, Ahmed Alsaedi, Areej S. Aljahdali, Sotiris K. Ntouyas, A study of coupled nonlinear generalized fractional differential equations with coupled nonlocal multipoint Riemann-Stieltjes and generalized fractional integral boundary conditions, 2023, 9, 2473-6988, 1576, 10.3934/math.2024078
    16. Bakhtawar Pervaiz, Akbar Zada, Ioan‐Lucian Popa, Sana Ben Moussa, Hala H. Abd El‐Gawad, Analysis of fractional integro causal evolution impulsive systems on time scales, 2023, 46, 0170-4214, 15226, 10.1002/mma.9374
    17. Chengdai Huang, Lei Fu, Shuang Liu, Jinde Cao, Mahmoud Abdel‐Aty, Heng Liu, Dynamical bifurcations in a delayed fractional‐order neural network involving neutral terms, 2024, 0170-4214, 10.1002/mma.10434
    18. Leyla Soudani, Abdelkader Amara, Khaled Zennir, Junaid Ahmad, Duality of fractional derivatives: On a hybrid and non-hybrid inclusion problem, 2024, 0928-0219, 10.1515/jiip-2023-0098
    19. Alireza Khabiri, Ali Asgari, Reza Taghipour, Mohsen Bozorgnasab, Ahmad Aftabi-Sani, Hossein Jafari, Analysis of fractional Euler-Bernoulli bending beams using Green’s function method, 2024, 106, 11100168, 312, 10.1016/j.aej.2024.07.023
    20. Hassen Arfaoui, Polynomial decay of a linear system of PDEs via Caputo fractional‐time derivative, 2024, 47, 0170-4214, 10490, 10.1002/mma.10135
    21. Lai Van Phut, Finite-time stability analysis of fractional fuzzy differential equations with time-varying delay involving the generalized Caputo fractional derivative, 2024, 35, 1012-9405, 10.1007/s13370-024-01201-9
    22. Najat Almutairi, Sayed Saber, Hijaz Ahmad, The fractal-fractional Atangana-Baleanu operator for pneumonia disease: stability, statistical and numerical analyses, 2023, 8, 2473-6988, 29382, 10.3934/math.20231504
    23. Yuequn Gao, Ning Li, Fractional order PD control of the Hopf bifurcation of HBV viral systems with multiple time delays, 2023, 83, 11100168, 1, 10.1016/j.aej.2023.10.032
    24. Saud Fahad Aldosary, Mohamed M. A. Metwali, Manochehr Kazemi, Ateq Alsaadi, On integrable and approximate solutions for Hadamard fractional quadratic integral equations, 2024, 9, 2473-6988, 5746, 10.3934/math.2024279
    25. Farzaneh Safari, Juan J. Nieto, Numerical analysis with a class of trigonometric functions for nonlinear time fractional Wu-Zhang system, 2024, 86, 11100168, 194, 10.1016/j.aej.2023.11.065
    26. Shabir Ahmad, Aman Ullah, Abd Ullah, Ngo Van Hoa, Fuzzy natural transform method for solving fuzzy differential equations, 2023, 27, 1432-7643, 8611, 10.1007/s00500-023-08194-w
    27. Zhiyao Ma, Ke Sun, Shaocheng Tong, Adaptive asymptotic tracking control of uncertain fractional-order nonlinear systems with unknown control coefficients and actuator faults, 2024, 182, 09600779, 114737, 10.1016/j.chaos.2024.114737
    28. B. Radhakrishnan, T. Sathya, M. A. Alqudah, W. Shatanawi, T. Abdeljawad, Existence Results for Nonlinear Hilfer Pantograph Fractional Integrodifferential Equations, 2024, 23, 1575-5460, 10.1007/s12346-024-01069-x
    29. Mohammad Tavazoei, Mohammad Saleh Tavazoei, Fractional systems with commensurate orders inherit the monotonicity of magnitude-frequency response from their integer-order counterparts, 2024, 1077-5463, 10.1177/10775463241303209
    30. Prakash Raj Murugadoss, Venkatesh Ambalarajan, Vinoth Sivakumar, Prasantha Bharathi Dhandapani, Dumitru Baleanu, Analysis of Dengue Transmission Dynamic Model by Stability and Hopf Bifurcation with Two-Time Delays, 2023, 28, 2768-6701, 10.31083/j.fbl2806117
    31. Shazia Sadiq, Mujeeb ur Rehman, Solution of fractional Sturm–Liouville problems by generalized polynomials, 2024, 0264-4401, 10.1108/EC-04-2024-0356
    32. Mingfang Lin, Zhonghui Ou, The analytical method of two-term time-fractional advection–dispersion–reaction models with sorption process, 2025, 114, 11100168, 702, 10.1016/j.aej.2024.11.112
    33. Mohamed Rhaima, Lassaad Mchiri, Abdellatif Ben Makhlouf, Existence, Uniqueness, and Averaging Principle for a Class of Fractional Neutral Itô–Doob Stochastic Differential Equations, 2025, 0170-4214, 10.1002/mma.10850
    34. Xin Chen, Bang‐Bang and Linear Quadratic Zero‐Sum Game for Noncausal and Causal Systems, 2025, 0170-4214, 10.1002/mma.10892
    35. Md. Samshad Hussain Ansari, Muslim Malik, Controllability of multi-term fractional-order impulsive dynamical systems with φ
    -Caputo fractional derivative, 2025, 1311-0454, 10.1007/s13540-025-00393-6
  • 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(3376) PDF downloads(371) Cited by(1)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog