We discuss the existence and uniqueness of solutions for a new class of sequential q-fractional integrodifferential equations with q-antiperiodic boundary conditions. Our results rely on the standard tools of fixed-point theory such as Krasnoselskii's fixed-point theorem, Leray-Schauder nonlinear alternative, and Banach's contraction principle. An illustrative example is also presented.
1. Introduction
We consider aq-antiperiodic boundary value problem of sequentialq-fractional integrodifferential equations given by
(1)Dcqα(Dcqγ+λ)x(t)=Af(t,x(t))+BIqρg(t,x(t)),nnnnnnnnnnnnnnnn.nnnnnl0≤t≤1,0<q<1,x(0)=-x(1),(t(1-γ)Dqx(t))|t=0=-Dqx(1),
where Dcqαand Dcqγdenote the fractionalq-derivative of the Caputo type, 0<α,γ≤1, Iqρ(·)denotes Riemann-Liouville integral with0<ρ<1,f,gbeing given continuous functions,λ∈ℝ andA,Bbeing real constants.
The aim of the present study is to establish some existence and uniqueness results for the problem (1) by means of Krasnoselskii’s fixed-point theorem, Leray-Schauder nonlinear alternative, and Banach’s contraction principle. Though the tools employed in this work are standard, yet their exposition in the framework of the given problem is new.
Fractional calculus has developed into a popular mathematical modelling tool for many real world phenomena occurring in physical and technical sciences, see, for example, [1–4]. A fractional-order differential operator distinguishes itself from an integer-order differential operator in the sense that it is nonlocal in nature and can describe the memory and hereditary properties of some important and useful materials and processes. This feature has fascinated many researchers and several results ranging from theoretical analysis to asymptotic behavior and numerical methods for fractional differential equations have been established. For some recent work on the topic, see [5–12] and references therein.
The mathematical modeling of linear control systems, concerning the controllability of systems consisting of a set of well-defined interconnected objects, is based on the linear systems of divided difference functional equations. The controllability in mathematical control theory studies the concepts such as controllability of the state, controllability of the output, controllability at the origin, and complete controllability. Theq-difference equations play a key role in the control theory as these equations are always completely controllable and appear in theq-optimal control problem [13]. The variationalq-calculus is known as a generalization of the continuous variational calculus due to the presence of an extra-parameterqwhose nature may be physical or economical. The study of theq-uniform lattice rely on theq-Euler equations. In other words, it suffices to solve theq-Euler-Lagrange equation for finding the extremum of the functional involved instead of solving the Euler-Lagrange equation [14]. One can find more details in a series of papers [15–21].
The subject of fractionalq-difference (q-fractional) equations is regarded as fractional analogue ofq-difference equations and has recently gained a considerable attention. For examples and details, we refer the reader to the works [22–33] and references therein while some earlier work on the subject can be found in [34–36]. The present work is motivated by recent interest in the study of fractional-order differential equations.
2. Preliminaries on Fractional q-Calculus
Let us describe the notations and terminology forq-fractional calculus [35].
For a real parameterq∈ℝ+∖{1}, aq-real number denoted by[a]qis defined by
(2)[a]q=1-qa1-q,a∈ℝ.
Theq-analogue of the Pochhammer symbol (q-shifted factorial) is defined as
(3)(a;q)0=1,(a;q)k=∏i=0k-1(1-aqi),k∈ℕ∪{∞}.
Theq-analogue of the exponent (x-y)kis
(4)(x-y)(0)=1,(x-y)(k)=∏j=0k-1(x-yqj),nnnnnnnnnnnn.nnnn.n.nnnnnk∈ℕ,x,y∈ℝ.
Theq-gamma functionΓq(y)is defined as
(5)Γq(y)=(1-q)(y-1)(1-q)y-1,
where y∈ℝ∖{0,-1,-2,…}. Observe that Γq(y+1)=[y]qΓq(y).
Definition 1 (see [35]).
Letfbe a function defined on[0,1].The fractional q-integral of the Riemann-Liouville type of orderβ≥0is(Iq0f)(t)=f(t)and
(6)Iqβf(t)≔∫0t(t-qs)(β-1)Γq(β)f(s)dqs=tβ(1-q)β∑k=0∞qk(qβ;q)k(q;q)kf(tqk),nnnnnnnnnnnnnn.nβ>0,t∈[0,1].
Observe that the aboveq-integral reduces to the following one forβ=1.
(7)Iqf(t)∶=∫0tf(s)dqs=t(1-q)∑k=0∞qkf(tqk).
Further details ofq-integrals and fractionalq-integrals can be found respectively in Section 1.3 and Section 4.2 of the text [35].
Remark 2.
The semigroup property holds forq-fractional integration (Proposition 4.3 [35]):
(8)IqγIqβf(t)=Iqβ+γf(t);γ,β∈ℝ+.
Further, it has been shown in Lemma 6 of [37] that
(9)Iqβ(x)(υ)=Γq(υ+1)Γq(β+υ+1)(x)(β+υ),0<x<a,β∈ℝ+,υ∈(-1,∞).
Before giving the definition of fractionalq-derivative, we recall the concept ofq-derivative.
Letfbe a real valued function defined on aq-geometric setA(|q|≠1). Then theq-derivative of a function fis defined as
(10)Dqf(t)=f(t)-f(qt)t-qt,t∈A∖{0}.
For0∈A, theq-derivative at zero is defined for|q|<1by
(11)Dqf(0)=limn→∞f(tqn)-f(0)tqn,t∈A∖{0}.
Provided that the limit exists and does not depend ont.
The Caputo fractionalq-derivative of orderβ>0is defined by
(13)cDqβf(t)=Iq⌈β⌉-βDq⌈β⌉f(t),
where ⌈β⌉is the smallest integer greater than or equal toβ.
Next we enlist some properties involving Riemann-Liouvilleq-fractional integral and Caputo fractionalq-derivative (Theorem 5.2 [35]):
(14)IqβDcqβf(t)=f(t)-∑k=0⌈β⌉-1tkΓq(k+1)(Dqkf)(0+),.nnnnnnnnnnnnnnnnnnnnnn∀t∈(0,a],β>0;cDqβIqβf(t)=f(t),∀t∈(0,a],β>0.
Now we establish a lemma that plays a key role in the sequel.
Lemma 4.
For a givenh∈C([0,1],ℝ), the boundary value problem
(15)Dcqα(Dqγc+λ)x(t)=h(t),0≤t≤1,0<q<1,x(0)=-x(1),(t(1-γ)Dqx(t))|t=0=-Dqx(1)
is equivalent to theq-integral equation
(16)x(t)=∫0t(t-qu)(γ-1)Γq(γ)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu+(1-2tγ)4[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu-12∫01(1-qu)(γ-1)Γq(γ)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu.
Proof.
It is well known that the solution ofq-fractional equation in (15) can be written as
(17)x(t)=∫0t(t-qu)(γ-1)Γq(γ)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu-tγΓq(γ+1)c0-c1.
Differentiating (17), we obtain
(18)Dqx(t)=∫0t(t-qu)(γ-2)Γq(γ-1)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu-[γ]qtγ-1Γq(γ+1)c0.
Using the boundary conditions (15) in (17) and (18) and solving the resulting expressions forc0and c1, we get
(19)c0=Γq(γ+1)2[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu,c1=12∫01(1-qu)(γ-1)Γq(γ)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu-14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×(∫0u(u-qm)(α-1)Γq(α)h(m)dqm-λx(u))dqu.
Substituting the values ofc0andc1in (17) yields the solution (16). The converse follows in a straightforward manner. This completes the proof.
Let 𝒞=C([0,1],ℝ) denote the Banach space of all continuous functions from[0,1]into ℝ endowed with the usual norm defined by∥x∥=sup{|x(t)|,t∈[0,1]}.
In view of Lemma 4, we define an operator 𝒰:𝒞→𝒞 as
(20)(𝒰x)(t)=∫0t(t-qu)(γ-1)Γq(γ)×(A∫0u(u-qm)(α-1)Γq(α)n×f(m,x(m))dqm+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×g(m,x(m))dqm∫0u(u-qm)(α-1)Γq(α)20000-λx(u)∫0u(u-qm)(α-1)Γq(α))dqu+(1-2tγ)4[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×(A∫0u(u-qm)(α-1)Γq(α)nll×f(m,x(m))dqm+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)∫0u(u-qm)(α-1)Γq(α)×g(m,x(m))dqm-λx(u))dqu-12∫01(1-qu)(γ-1)Γq(γ)×(A∫0u(u-qm)(α-1)Γq(α)×f(m,x(m))dqm+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×g(m,x(m))dqm∫0u(u-qm)(α-1)Γq(α)-λx(u))dqu.
Observe that the problem (1) has solutions only if the operator equationx=𝒰xhas fixed points.
3. Main Results
For the forthcoming analysis, the following conditions are assumed.
f,g:[0,1]×ℝ→ℝ are continuous functions such that|f(t,x)-f(t,y)|≤L1|x-y|and|g(t,x)-g(t,y)|≤L2|x-y|, for all t∈[0,1],L1,L2>0,x,y∈ℝ.
There existζ1,ζ2∈C([0,1],ℝ+)with|f(t,x)|≤ζ1(t),|g(t,x)|≤ζ2(t), for all (t,x)∈[0,1]×ℝ, where∥ζi∥=supt∈[0,1]|ζi(t)|,i=1,2.
For computational convenience, we set
(21)δ1=32Γq(α+γ+1)+14[γ]qΓq(α+γ),δ2=32Γq(α+ρ+γ+1)+14[γ]qΓq(α+ρ+γ),δ3=32Γq(γ+1)+14Γq(γ+1),(22)Λ=L[|A|(14[γ]qΓq(α+γ)+12Γq(α+γ+1))+|B|(14[γ]qΓq(α+ρ+γ)+12Γq(α+ρ+γ+1))]+|λ|[14Γq(γ+1)+12Γq(γ+1)].
Our first existence result is based on Krasnoselskii’s fixed point theorem.
Lemma 5 (see, Krasnoselskii [38]).
LetYbe a closed, convex, bounded, and nonempty subset of a Banach spaceX.LetQ1,Q2be the operators such that (i)Q1x+Q2y∈Ywheneverx,y∈Y; (ii)Q1is compact and continuous; and (iii)Q2is a contraction mapping. Then there existsz∈Ysuch thatz=Q1z+Q2z.
Theorem 6.
Letf,g:[0,1]×ℝ→ℝ be continuous functions satisfying (A1)–(A2).FurthermoreΛ<1, whereΛis given by (22) andL=max{L1,L2}.Then the problem (1) has at least one solution on[0,1].
Proof.
Consider the setBσ={x∈𝒞:∥x∥≤σ}, whereσis given by
(23)σ≥|A|∥ζ1∥δ1+|B|∥ζ2∥δ21-|λ|δ3,1-|λ|δ3>0.
Define operators𝒰1and𝒰2onBσas
(24)(𝒰1x)(t)=∫0t(t-qu)(γ-1)Γq(γ)×(A∫0u(u-qm)(α-1)Γq(α)f(m,x(m))dqmnnnnn+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)nnnnnnnnnn×g(m,x(m))dqmnnnnnnnnnl∫0u(u-qm)(α-1)Γq(α)-λx(u))dqu,t∈[0,1],(𝒰2x)(t)=(1-2tγ)4[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)nnnn×(A∫0u(u-qm)(α-1)Γq(α)∫0u(u-qm)(α+ρ-1)Γq(α+ρ)nnnnnnn×f(m,x(m))dqmnnnnnn+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)nnnnnnlnnnn×g(m,x(m))dqmnnnnnnnnnnnnnnn∫0u(u-qm)(α+ρ-1)Γq(α+ρ)nnnnnn-λx(u))dqu-12∫01(1-qu)(γ-1)Γq(γ)×(A∫0u(u-qm)(α-1)Γq(α)f(m,x(m))dqm+B∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×g(m,x(m))dqm∫0u(u-qm)(α-1)Γq(α)-λx(u))dqu,t∈[0,1].
Forx,y∈Bσ, we find that
(25)∥𝒰1x+𝒰2y∥≤|A|∥ζ1∥δ1+|B|∥ζ2∥δ2+|λ|σδ3≤σ.
Thus,𝒰1x+𝒰2y∈Bσ.Continuity offandgimply that the operator𝒰1 is continuous. Also,𝒰1is uniformly bounded onBσas
(26)∥𝒰1x∥≤|A|∥ζ1∥Γq(α+γ+1)+|B|∥ζ2∥Γq(α+ρ+γ+1)+|λ|σΓq(γ+1).
Now, we prove the compactness of the operator𝒰1.In view of (A1), we define
(27)sup(t,x)∈[0,1]×Bσ|f(t,x)|=f¯,sup(t,x)∈[0,1]×Bσ|g(t,x)|=g¯.
Consequently, fort1,t2∈[0,1], we have
(28)∥(𝒰1x)(t2)-(𝒰1x)(t1)∥≤∫0t1(t2-qu)(γ-1)-(t1-qu)(γ-1)Γq(γ)×[|A|f¯∫0u(u-qm)(α-1)Γq(α)dqm+|B|g¯×∫0u(u-qm)(α+ρ-1)Γq(α+ρ)dqm+|λ|σ]dqu+∫t1t2(t2-qu)(γ-1)Γq(γ)×[|A|f¯∫0u(u-qm)(α-1)Γq(α)dqm+|B|g¯×∫0u(u-qm)(α+ρ-1)Γq(α+ρ)dqm+|λ|σ]dqu,
which is independent ofxand tends to zero ast2→t1. Thus,𝒰1is relatively compact onBσ. Hence, by the Arzelá-Ascoli Theorem,𝒰1is compact onBσ.Now, we shall show that𝒰2is a contraction.
From (A1)and forx,y∈Bσ, we have
(29)∥𝒰2x-𝒰2y∥≤supt∈[0,1]{14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)nn.×|f(m,x(m))-f(m,y(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))-g(m,y(m))|dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)-y(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))-f(m,y(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))-g(m,y(m))|dqm+∫0u(u-qm)(α-1)Γq(α)|λ||x(u)-y(u)|]dqu}≤supt∈[0,1]{14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)L1×|x(m)-y(m)|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)nn.×L2|x(m)-y(m)|dqm+∫01(1-qu)(γ-2)Γq(γ-1)|λ||x(u)-y(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)n.l×L1|x(m)-y(m)|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×L2|x(m)-y(m)|dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)-y(u)|]dqu}≤[L[|A|(14[γ]qΓq(α+γ)+12Γq(α+γ+1))+|B|(14[γ]qΓq(α+ρ+γ)+12Γq(α+ρ+γ+1))]+|λ|[14[γ]qΓq(γ)+12Γq(γ+1)]]∥x-y∥=Λ∥x-y∥,
where we have used (22). In view of the assumptionΛ<1, the operator𝒰2is a contraction. Thus, all the conditions of Lemma 5 are satisfied. Hence, by the conclusion of Lemma 5, the problem (1) has at least one solution on[0,1].
Our next result is based on Leray-Schauder nonlinear alternative.
Lemma 7 (nonlinear alternative for single valued maps, see [39]).
LetEbe a Banach space,Ca closed, convex subset of E,Wan open subset ofC, and0∈W.Suppose that 𝒰:W¯→Cis a continuous, compact (i.e.,𝒰(W¯)is a relatively compact subset ofC) map. Then either
𝒰 has a fixed point in W¯, or
there is ax∈∂W(the boundary ofWinC) andκ∈(0,1)withx=κ𝒰(x).
Theorem 8.
Letf,g:[0,1]×ℝ→ℝ be continuous functions and the following assumptions hold:
there exist functionsν1,ν2∈C([0,1],ℝ+), and nondecreasing functionsϑ1,ϑ2:ℝ+→ℝ+such that|f(t,x)|≤ν1(t)ϑ1(∥x∥),|g(t,x)|≤ν2(t)ϑ2(∥x∥), for all (t,x)∈[0,1]×ℝ;
there exists a constantω>0such that
(30)ω>|A|∥ν1∥ϑ1(ω)δ1+|B|∥ν2∥ϑ2(ω)δ21-|λ|δ3,1-|λ|δ3>0.
Then the boundary value problem (1) has at least one solution on[0,1].
Proof.
Consider the operator 𝒰:𝒞→𝒞 defined by (20). The proof consists of several steps.
It is easy to show that 𝒰 is continuous.
𝒰 maps bounded sets into bounded sets in C([0,1]×ℝ).
For a positive numberr¯, letBr¯={x∈𝒞:∥x∥≤r¯} be a bounded set in C([0,1]×ℝ) andx∈Br¯. Then, we have
(31)∥(𝒰x)∥≤supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))|dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu+14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))|dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))|dqm∫0t(t-qu)(γ-1)Γq(γ)+|λ||x(u)|]dqu}≤supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)ν1(m)ϑ1(∥x∥)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×ν2(m)ϑ2(∥x∥)dqm∫0t(t-qu)(γ-1)Γq(γ)+|λ||x(u)|]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)×ν1(m)ϑ1(∥x∥)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×ν2(m)ϑ2(∥x∥)dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×ν1(m)ϑ1(∥x∥)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×ν2(m)ϑ2(∥x∥)dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu}≤|A|∥ν1∥ϑ1(∥x∥)×{∫01(1-qu)(γ-1)Γq(γ)[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu}+|B|∥ν2∥ϑ2(∥x∥)×{∫01(1-qu)(γ-1)Γq(γ)[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+12∫01(1-qu)(γ-1)Γq(γ)nn.×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu}+|λ|∥x∥{∫01(1-qu)(γ-1)Γq(γ)dqu+14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)dqu+12∫01(1-qu)(γ-1)Γq(γ)dqu}≤|A|∥ν1∥ϑ1(r¯)δ1+|B|∥ν2∥ϑ2(r¯)δ2+|λ|r¯δ3≤r¯.
This shows that 𝒰x∈Br¯.
𝒰 maps bounded sets into equicontinuous sets of C([0,1]×ℝ).
Let t1,t2∈[0,1] with t1<t2 and x∈Br¯, where Br¯is a bounded set of C([0,1],ℝ). Then, we obtain
(32)∥(𝒰x)(t2)-(𝒰x)(t1)∥≤|∫0t1(t2-qu)(γ-1)-(t1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)ν1(m)ϑ1(r¯)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×∫0u(u-qm)(α+ρ-1)Γq(α+ρ)ν2(m)ϑ2(r¯)dqm+|λ|r¯]dqu+∫t1t2(t2-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)ν1(m)ϑ1(r¯)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×ν2(m)ϑ2(r¯)dqm∫0u(u-qm)(α-1)Γq(α)+|λ|r¯]dqu|+(t2γ-t1γ)2[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)ν1(m)ϑ1(r¯)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×ν2(m)ϑ2(r¯)dqm∫0u(u-qm)(α-1)Γq(α)+|λ|r¯]dqu.
Obviously the right-hand side of the above inequality tends to zero independently ofx∈Br¯ as t2-t1→0.Therefore, it follows by the Arzelá-Ascoli theorem that 𝒰:𝒞→𝒞 is completely continuous.
Letxbe a solution of the given problem such thatx=κ𝒰xforκ∈(0,1).Then, fort∈[0,1], it follows by the procedure used to establish (ii) that
(33)|x(t)|=|κ(𝒰x)(t)|≤|A|∥ν1∥ϑ1(∥x∥)δ1+|B|∥ν2∥ϑ2(∥x∥)δ2+|λ|∥x∥δ3.
Consequently, we have
(34)∥x∥≤|A|∥ν1∥ϑ1(∥x∥)δ1+|B|∥ν2∥ϑ2(∥x∥)δ21-|λ|δ3.
In view of(A4), there existsωsuch that∥x∥≠ω. Let us set
(35)W={x∈𝒞:∥x∥<ω}.
Note that the operator 𝒰:W¯→C([0,1],ℝ)is continuous and completely continuous. From the choice ofW, there is nox∈∂Wsuch thatx=κ𝒰(x)for someκ∈(0,1). In consequence, by the nonlinear alternative of Leray-Schauder type (Lemma 7), we deduce that 𝒰has a fixed pointx∈W¯which is a solution of the problem (1). This completes the proof.
Finally we show the existence of a unique solution of the given problem by applying Banach’s contraction mapping principle (Banach fixed-point theorem).
Theorem 9.
Suppose that the assumption (A1) holds and
(36)Λ¯=(LΛ1+|λ|δ3)<1,Λ1=|A|δ1+|B|δ2,
whereδ1,δ2, andδ3are given by (21) andL=max{L1,L2}. Then the boundary value problem (1) has a unique solution.
Proof.
Fix M=max{M1,M2}, where M1,M2are finite numbers given byM1=supt∈[0,1]|f(t,0)|,M2=supt∈[0,1]|g(t,0)|. Selecting σ≥MΛ1/(1-Λ¯), we show that 𝒰Bσ⊂Bσ, where Bσ={x∈𝒞:∥x∥≤σ}. For x∈Bσ, we have
(37)∥(𝒰x)∥≤supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))|dqm∫0t(t-qu)(γ-1)Γq(γ)+|λ||x(u)|]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))|dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×∫0u(u-qm)(α+ρ-1)Γq(α+ρ)|g(m,x(m))|dqm+|λ||x(u)||A|∫0u(u-qm)(α-1)Γq(α)]dqu∫0t(t-qu)(γ-1)Γq(γ)}≤supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×(|f(m,x(m))-f(m,0)|+|f(m,0)|)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×(|g(m,x(m))-g(m,0)|+|g(m,0)|)dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)×(|f(m,x(m))-f(m,0)|+|f(m,0)|)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×(|g(m,x(m))-g(m,0)|+|g(m,0)|)dqm+∫0u(u-qm)(α-1)Γq(α)|λ||x(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×(|f(m,x(m))-f(m,0)|+|f(m,0)|)dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×(|g(m,x(m))-g(m,0)|+|g(m,0)|)dqm∫0u(u-qm)(α-1)Γq(α)+|λ||x(u)|]dqu}≤|A|(L1σ+M1)×supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[∫0u(u-qm)(α-1)Γq(α)dqm]dqu}+|B|(L2σ+M2)×supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[∫0u(u-qm)(α+ρ-1)Γq(α+ρ)dqm]dqu+14[γ]q∫01(1-qu)(γ-2)Γq(γ-1)×[∫0u(u-qm)(α+ρ-1)Γq(α+ρ)dqm]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[∫0u(u-qm)(α+ρ-1)Γq(α+ρ)dqm]dqu}+|λ|σsupt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)dqu+14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)dqu+12∫01(1-qu)(γ-1)Γq(γ)dqu}≤MΛ1+Λ¯σ≤σ
This shows that 𝒰Bσ⊂Bσ.For x,y∈ℝ, we obtain
(38)∥𝒰x-𝒰y∥≤supt∈[0,1]{∫0t(t-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))-f(m,y(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))-g(m,y(m))|dqm+∫0u(u-qm)(α-1)Γq(α)|λ||x(u)-y(u)|]dqu+14[γ]q×∫01(1-qu)(γ-2)Γq(γ-1)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))-f(m,y(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))-g(m,y(m))|dqm+∫0u(u-qm)(α-1)Γq(α)|λ||x(u)-y(u)|]dqu+12∫01(1-qu)(γ-1)Γq(γ)×[|A|∫0u(u-qm)(α-1)Γq(α)×|f(m,x(m))-f(m,y(m))|dqm+|B|∫0u(u-qm)(α+ρ-1)Γq(α+ρ)×|g(m,x(m))-g(m,y(m))|dqm∫0t(t-qu)(γ-1)Γq(γ)+∫0u(u-qm)(α-1)Γq(α)|λ||x(u)-y(u)|]dqu}≤Λ¯∥x-y∥.
SinceΛ¯∈(0,1)by the given assumption, therefore 𝒰 is a contraction. Hence, it follows by Banach’s contraction principle that the problem (1) has a unique solution.
Example 10.
Consider aq-fractional integrodifferential equation withq-antiperiodic boundary conditions given by
(39)Dcqα(Dqγc+115)x(t)=15f(t,x(t))+18Iqρg(t,x(t)),nnnnnnnnnnnnnn.0<t<1,0<q<1,x(0)=-x(1),(t(1-γ)Dqx(t))|t=0=-Dqx(1),
where α=γ=ρ=q=1/2,λ=1/15, A=1/5,B=1/8,f(t,x)=(t/20π)sin2πx+(t+1)(1+x)4/(1+(1+x)4), g(t,x)=(1/8π)tan-1x+1/16.With the given data,δ1≃1.92678,δ2≃1.72332,δ3≃1.90037,and
(40)|f(t,x)|≤(t+1)(∥x∥10+1),|g(t,x)|≤18.
Clearly,ν1(t)=t+1,ϑ1(∥x∥)=∥x∥/10+1,ν2(t)=1/8,ϑ1(∥x∥)=1, and the condition (A4)implies that ω>1.00176. Thus all the assumptions of Theorem 8 are satisfied. Hence, the conclusion of Theorem 8 applies to the problem (39).
Example 11.
Consider the followingq-fractionalq-antiperiodic boundary value problem:
(41)Dcqα(Dqγc+18)x(t)=12f(t,x(t))+14Iqρg(t,x(t)),nnnnnnnnnnn0<t<1,0<q<1,x(0)=-x(1),(t(1-γ)Dqx(t))|t=0=-Dqx(1),
where α=γ=ρ=q=A=1/2, λ=1/8,B=1/4,f(t,x)=(1/(t2+2))(|x|/(1+|x|))+sin2t,g(t,x)=(1/4)tan-1x+cos2t+t3+5.With the given data, it is found thatL1=1/2,L2=1/4as |f(t,x)-f(t,y)|≤(1/2)|x-y|, |g(t,x)-g(t,y)|≤(1/4)|x-y|.Clearly L=max{L1,L2}=1/2. Moreover, δ1≃1.92678,δ2≃1.72332,and δ3≃1.90037. Using the given values, it is found thatΛ¯≃0.934655<1. Thus all the assumptions of Theorem 9 are satisfied. Hence, by the conclusion of Theorem 9, there exists a unique solution for the problem (41).
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgment
This work was partially supported by Deanship of Scientific Research (DSR), King Abdulaziz University, Jeddah, Saudi Arabia.
SamkoS. G.KilbasA. A.MarichevO. I.1993Yverdon, SwitzerlandGordon and Breach Sciencexxxvi+976MR1347689PodlubnyI.1999198San Diego, Calif, USAAcademic Pressxxiv+340Mathematics in Science and EngineeringMR1658022KilbasA. A.SrivastavaH. M.TrujilloJ. J.2006204Amsterdam, The NetherlandsElsevier Science B.V.xvi+523North-Holland Mathematics StudiesMR2218073BaleanuD.DiethelmK.ScalasE.TrujilloJ. J.20123Boston, Mass, USAWorld Scientificxxiv+400Series on Complexity, Nonlinearity and Chaos10.1142/9789814355216MR2894576BăleanuD.MustafaO. G.AgarwalR. P.On Lp-solutions for a class of sequential fractional differential equations201121852074208110.1016/j.amc.2011.07.024MR2831483LiuX.JiaM.GeW.Multiple solutions of a p-Laplacian model involving a fractional derivative20132013, article 12610.1186/1687-1847-2013-126MR3064004AgarwalR. P.AhmadB.Existence theory for anti-periodic boundary value problems of fractional differential equations and inclusions20116231200121410.1016/j.camwa.2011.03.001MR2824708ZBL1228.34009FordN. J.MorgadoM. L.Fractional boundary value problems: analysis and numerical methods201114455456710.2478/s13540-011-0034-4MR2846376ZBL1273.65098AghajaniA.JalilianY.TrujilloJ. J.On the existence of solutions of fractional integro-differential equations2012151446910.2478/s13540-012-0005-4MR2872111ZBL1279.45008O'ReganD.StaněkS.Fractional boundary value problems with singularities in space variables201371464165210.1007/s11071-012-0443-xMR3030127ZBL1268.34023AhmadB.NietoJ. J.Boundary value problems for a class of sequential integrodifferential equations of fractional order20132013810.1155/2013/149659149659MR3046762ZBL1267.45018ZhangL.AhmadB.WangG.AgarwalR. P.Nonlinear fractional integro-differential equations on unbounded domains in a Banach space2013249515610.1016/j.cam.2013.02.010MR3037805BangerezakoG.q-difference linear control systems20111791229124910.1080/10236190902899115MR2825244ZBL1234.39001BangerezakoG.Variational q-calculus2004289265066510.1016/j.jmaa.2003.09.004MR2026931LoganJ. D.First integrals in the discrete variational calculus19739210220MR032839710.1007/BF01832628ZBL0268.49022AgrawalO. P.Some generalized fractional calculus operators and their applications in integral equations201215470071110.2478/s13540-012-0047-7MR2974327GuermahS.DjennouneS.BettayebM.Controllability and observability of linear discrete-time fractional-order systems200818221322210.2478/v10006-008-0019-6MR2426575ZBL1234.93014BartosiewiczZ.PawłuszewiczE.Realizations of linear control systems on time scales2006354769786MR2323260ZBL1133.93033MozyrskaD.BartosiewiczZ.On observability concepts for nonlinear discrete-time fractional order control systems20104305312AbdeljawadT.JaradF.BaleanuD.Variational optimal-control problems with delayed arguments on time scales2009200910.1155/2009/840386840386MR2577651ZBL1184.49023MainardiF.CarpinteriA.MainardiF.Fractional calculus: some basic problems in continuum and statistical mechanics1997New York, NY, USASpringer291348FerreiraR. A. C.Nontrivial solutions for fractional q-difference boundary value problems201070110MR2740675AhmadB.NietoJ. J.AlsaediA.Al-HutamiH.Existence of solutions for nonlinear fractional q-difference integral equations with two fractional orders and nonlocal four-point boundary conditionsvol. 351, no. 5, pp. 2890-2909, 201410.1016/j.jfranklin.2014.01.020GoodrichC. S.Existence and uniqueness of solutions to a fractional difference equation with nonlocal conditions201161219120210.1016/j.camwa.2010.10.041MR2754129ZBL1211.39002MaJ.YangJ.Existence of solutions for multi-point boundary value problem of fractional q-difference equation201192110MR2861421GraefJ. R.KongL.Positive solutions for a class of higher order boundary value problems with fractional q-derivatives2012218199682968910.1016/j.amc.2012.03.006MR2916148AhmadB.NtouyasS. K.Existence of solutions for nonlinear fractional q-difference inclusions with nonlocal Robin (separated) conditions20131031333135110.1007/s00009-013-0258-0MR3080212LiX.HanZ.SunS.Existence of positive solutions of nonlinear fractional q-difference equation with parameter20132013, article 26010.1186/1687-1847-2013-260MR3110765AlsaediA.AhmadB.Al-HutamiH.A study of nonlinear fractional q-difference equations with nonlocal integral boundary conditions20132013810.1155/2013/410505410505MR3121496JaradF.AbdeljawadT.BaleanuD.Stability of q-fractional non-autonomous systems201314178078410.1016/j.nonrwa.2012.08.001MR2969872AbdeljawadT.BaleanuD.JaradF.AgarwalR. P.Fractional sums and differences with binomial coefficients20132013610.1155/2013/104173104173MR3063724ZBL1264.26006FerreiraR. A. C.Positive solutions for a class of boundary value problems with fractional q-differences201161236737310.1016/j.camwa.2010.11.012MR2754144AshyralyevA.NalbantN.SzenY.Structure of fractional spaces generated by second order difference operators2014351271373110.1016/j.jfranklin.2013.07.009Al-SalamW. A.Some fractional q-integrals and q-derivatives196715135140MR021884810.1017/S0013091500011469AnnabyM. H.MansourZ. S.20122056Berlin, GermanySpringerxx+318Lecture Notes in Mathematics10.1007/978-3-642-30898-7MR2963764AgarwalR. P.Certain fractional q-integrals and q-derivatives196966365370MR0247389RajkovićP. M.MarinkovićS. D.StankovićM. S.On q-analogues of Caputo derivative and Mittag-Leffler function2007104359373MR2378985SmartD. R.1974Cambridge University Pressviii+93MR0467717GranasA.DugundjiJ.2003New York, NY, USASpringerxvi+690Springer Monographs in MathematicsMR1987179