Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Open Mathematics

formerly Central European Journal of Mathematics

Editor-in-Chief: Vespri, Vincenzo / Marano, Salvatore Angelo


IMPACT FACTOR 2018: 0.726
5-year IMPACT FACTOR: 0.869

CiteScore 2018: 0.90

SCImago Journal Rank (SJR) 2018: 0.323
Source Normalized Impact per Paper (SNIP) 2018: 0.821

Mathematical Citation Quotient (MCQ) 2018: 0.34

ICV 2018: 152.31

Open Access
Online
ISSN
2391-5455
See all formats and pricing
More options …
Volume 15, Issue 1

Issues

Volume 13 (2015)

Positive solutions for Hadamard differential systems with fractional integral conditions on an unbounded domain

Tariboon Jessada
  • Corresponding author
  • Nonlinear Dynamic Analysis Research Center. Department of Mathematics. Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok, Bangkok 10800, Thailand
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Sotiris K. Ntouyas
  • Department of Mathematics, University of Ioannina, 451 10 Ioannina, Ioannina, Greece
  • Nonlinear Analysis and Applied Mathematics (NAAM)-Research Group, Department of Mathematics, Faculty of Science, King Abdulaziz University, P.O. Box 80203, Jeddah 21589 Saudi Arabia
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Suphawat Asawasamrit
  • Nonlinear Dynamic Analysis Research Center. Department of Mathematics. Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok, Bangkok 10800, Thailand
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Chanon Promsakon
  • Nonlinear Dynamic Analysis Research Center. Department of Mathematics. Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok, Bangkok 10800, Thailand
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
Published Online: 2017-05-25 | DOI: https://doi.org/10.1515/math-2017-0057

Abstract

In this paper, we investigate the existence of positive solutions for Hadamard type fractional differential system with coupled nonlocal fractional integral boundary conditions on an infinite domain. Our analysis relies on Guo-Krasnoselskii’s and Leggett-Williams fixed point theorems. The obtained results are well illustrated with the aid of examples.

Keywords: Hadamard fractional differential systems; Nonlocal boundary conditions; Hadamard fractional integral conditions; Positive solutions; Fixed point theorems

MSC 2010: 26A33; 34A08; 34B18; 34B40

1 Introduction

Fractional calculus and fractional differential equations have been studied extensively during the last decades. Fractional derivatives provide a more excellent tool for the description of memory and hereditary properties of various materials and processes than integer derivatives. Engineers and scientists have developed new models that involve fractional differential equations. These models have been applied successfully in, e.g., physics, biomathematics, blood flow phenomena, ecology, environmental issues, viscoelasticity, aerodynamics, electrodynamics of complex medium, electrical circuits, electron-analytical chemistry, control theory, etc. For a systematic development of the topic, we refer to the books [1]-[7]. As an important issue for the theory of fractional differential equations, the existence of solutions to kinds of boundary value problems has attracted many scholars attention, and lots of excellent results have been obtained by means of fixed point theorems, upper and lower solutions technique, and so forth. A variety of results on initial and boundary value problems of fractional differential equations and inclusions can easily be found in the literature on the topic. For some recent results, we can refer to [8]-[18] and references cited therein.

Boundary value problems on infinite intervals appear often in applied mathematics and physics. Due to the fact that an infinite interval is noncompact, the discussion about boundary value problem on the infinite intervals is more complicated. Results on the existence of solutions of boundary value problems on infinite intervals for differential, difference and integral equations may be found in the monographs [19,20]. For boundary value problems of fractional order on infinite intervals we refer to [21]-[25].

Many researchers have shown their interest in the study of systems of fractional differential equations. The motivation for those works stems from both the intensive development of the theory of fractional calculus itself and the applications. See for example [26]-[30] where systems for fractional differential equations were studied by using Banach contraction mapping principle and Schaefer’s fixed point theorem.

Recently in [31] we investigated the existence of positive solutions for fractional differential equations of Hadamard type, with integral boundary condition on infinite intervals Dαu(t)+a(t)f(u(t))=0,1<α2,t(1,),(1) u(1)=0,Dα1u()=i=1mλiIβiu(η),(2)where Dα denotes the Hadamard fractional derivative of order α, η ∈ (1, ∞) and Iβi is the Hadamard fractional integral of order βi > 0, i = 1, 2,... , m and λi ≥ 0, i = 1, 2,... , m are given constants. For some recent results on positive solutions of fractional differential equations we refer to [32]-[37] and references cited therein.

In [38] the existence of positive solutions were studied for the following fractional system of differential equations subject to the nonlocal Riemann-Liouville fractional integral boundary conditions of the formDpx(t)+f(t,x(t)),y(t))=0,1<p2,t(0,1),Dqy(t)+g(t,x(t)),y(t))=0,1<q2,t(0,1),x(0)=0,x(1)=i=1mαiIγiy(η),y(0)=0,y(1)=j=1nβiIμjx(ξ),(3) where DΦ are Riemann-Liouville fractional derivatives of orders ϕ{p,q},f,gC([0,1]×R+2,R+),Iϕ are Riemann-Liouville fractional integrals of order ϕ ∈ {γi, μj}, αi, βj > 0, i = 1,..., m, j = 1,..., n and the fixed constants 0 < η < ξ < 1.

In this paper we investigate the existence of positive solutions for the following fractional system of Hadamard differential equations subject to the fractional integral boundary conditions on an unbounded domain Dpx(t)+a(t)f(x(t)),y(t))=0,1<p2,t(1,),Dqy(t)+b(t)g(x(t)),y(t))=0,1<q2,t(1,),x(1)=0,Dp1x()=i=1mλiIαiy(η),y(1)=0,Dq1y()=j=1nσjIβjx(ξ),(4) where DΦ are Hadamard fractional derivatives of orders Φ ∈ {p, q} with lower limit 1,f,gC([1,]×R+2,R+), IΦ are Hadamard fractional integrals of order Φ ∈ {αi, βj} with lower limit 1, λi, σj > 0, i,..., m, j = 1,..., n.

Applying first the well-known Guo-Krasnoselskii’s fixed point theorem we obtain the existence of at least one positive solution. Next we prove the existence of at least three distinct nonnegative solutions by using Leggett-Williams fixed point theorem.

The rest of the paper is organized as follows: In section 2, we present some preliminaries and lemmas that will be used to prove our main results. We also obtain the corresponding Green’s function and some of its properties. The main results are formulated and proved in Section 3. Especially in Subsection 3.1 we prove the existence of at least one positive solution while, in Subsection 3.2, we prove the existence of at least three distinct nonnegative solutions.

Examples illustrating our results are presented in Section 4.

2 Background materials and preliminaries

In this section, we present some notations and definitions of Hadamard fractional calculus (see [4]) and present preliminary results needed in our proofs later.

Definition 2.1

([14]). The Hadamard fractional integral of order Φ with lower limit 1 for a function g : [1, ∞) → is defined asIΦg(t)=1Γ(q)1tlogtsΦ1g(s)sds,Φ>0,(5)provided the integral exists, where log(•) = loge(•).

Definition 2.2

([4]). The Hadamard derivative of fractional order Φ with lower limit 1 for a function g : [1, ∞) → ℝ ℝ is defined asDϕg(t)=1Γ(nϕ)tddtn1tlogtsnϕ1g(s)sds,n1<ϕ<n,(6)where n = [Φ] + 1, [Φ] denotes the integer part of the real number Φ.

Lemma 2.3

([4, Property 2.24]). If a, α, β > 0 thenDaαlogtaβ1(x)=Γ(β)Γ(βα)logxaβα1.(7)andIαlogtaβ1(x)=Γ(β)Γ(β+α)logxaβ+α1.(8)

Lemma 2.4

([4]). Let q < 0 and x ∈ C[1, ∞) ⋂ L1[1, ∞). Then the Hadamard fractional differential equation Dqx(t) = 0 has the solutionsx(t)=i=1nci(logt)qi,and the following formula holds:IqDqx(t)=x(t)+i=1nci(logt)qi,where ci ∈ ℝ, i = 1, 2,...., n, and n — 1 < q < n.

Lemma 2.5

Suppose that the functions u,v ∈ C([1, ∞),ℝ+) and 1 < p, q ≤ 2. Then the following systemDpx(t)+u(t)=0,t(1,),Dqy(t)+v(t)=0,t(1,),x(1)=0,Dp1x()=i=1mλiIαiy(η),y(1)=0,Dq1y()=j=1nσjIβjx(ξ),(9)can be written in the equivalent integral equations of the formx(t)=1Γ(p)1tlogtsp1u(s)dss+(logt)p1[Γ(q)Ω1u(s)dssΓ(q)Ωi=1mλiΓ(q+αi)1ηlogηsq+αi1u(s)dss+Λ1Ω1v(s)dssΛ1Ωj=1nσjΓ(p+βj)1ξlogξsp+βj1u(s)dss],(10)andy(t)=1Γ(q)1t(logts)q1v(s)dss+(logt)q1[Γ(p)Ω1v(s)dssΓ(p)Ωj=1nσjΓ(p+βj)1ξ(logξs)p+βj1u(s)dss+Λ2Ω1u(s)dssΛ2Ωi=1mλiΓ(q+αi)1η(logηs)q+αi1v(s)dss],(11)whereΩ:=Γ(p)Γ(q)Λ1Λ2>0,withΛ1:=i=1mλiΓ(q)Γ(q+αi)(logη)q+αi1andΛ2:=j=1nσjΓ(p)Γ(p+βj)(logξ)p+βj1.

Proof

By applying the Hadamard fractional integral of orders p and q to both sides of the first two equations of system (9), respectively, we obtain x(t)=1Γ(p)1tlogtsp1u(s)dss+c1(logt)p1+c2(logt)p2,y(t)=1Γ(q)1tlogtsq1v(s)dss+k1(logt)q1+k2(logt)q2, where c1, c2, k1, k2 ∈ ℝ.

The conditions of (9) that x(1) = 0 and y(1) = 0 imply c2 = 0 and k2 = 0. Hence x(t)=1Γ(p)1tlogtsp1u(s)dss+c1(logt)p1,(12) and y(t)=1Γ(q)1tlogtsq1v(s)dss+k1(logt)q1.(13)

From Lemma 2.3, we obtain Dp1x(t)=1tu(s)dss+c1Γ(p)andDq1y(t)=1tv(s)dss+k1Γ(q).

By applying the Hadamard fractional integral of orders βj and αi to (12) and (13), and also substitution t = ξ and t = η, respectively, we get Iβjx(ξ)=1Γ(p+βj)1ξlogξsp+βj1u(s)dss+c1Γ(p)Γ(p+βj)(logξ)p+βj1, and Iαiy(η)=1Γ(q+αi)1ηlogηsq+αi1v(s)dss+k1Γ(q)Γ(q+αi)(logη)q+αi1.

Using the conditions of (9) and solving the system of linear equations we find the constants c1 and k1 as c1=Γ(q)Ω1u(s)dssΓ(q)Ωi=1mλiΓ(q+αi)1ηlogηsq+αi1v(s)dss+Λ1Ω1v(s)dssΛ1Ωj=1nσjΓ(p+βj)1ξlogξsp+βj1u(s)dss, and k1=Γ(p)Ω1v(s)dssΓ(p)Ωj=1nσjΓ(p+βj)1ξlogξsp+βj1u(s)dss+Λ2Ω1u(s)dssΛ2Ωi=1mλiΓ(q+αi)1ηlogηsq+αi1v(s)dss.

Substituting the values of c1 and k1 in (12) and (13), we deduce the integral equations (10) and (11), respectively.

The converse follows by direct computation. This completes the proof.  □

Lemma 2.6

(Green’s function). The integral equations ((10)) and ((11)), in Lemma 2.5, can be expressed in the form of Green’s functions asx(t)=1G1(t,s)u(s)dss+1G2(t,s)v(s)dss,(14)y(t)=1G3(t,s)v(s)dss+1G4(t,s)u(s)dss,(15)where the Green’s functions Gi (t, s), i = 1, 2, 3, 4 are given byG1(t,s)=gp(t,s)+Λ1Ωj=1nσj(logt)p1Γ(p+βj)gβjp(ξ,s),G2(t,s)=Γ(q)Ωi=1mλi(logt)p1Γ(q+αi)gαiq(η,s),G3(t,s)=gq(t,s)+Λ2Ωi=1mλi(logt)q1Γ(q+αi)gαiq(η,s),G4(t,s)=Γ(p)Ωj=1nσj(logt)q1Γ(p+βj)gβjp(ξ,s),wheregϕ(t,s)=(logt)ϕ1(log(t/s))ϕ1I(ϕ),1st<,(logt)ϕ1Γ(ϕ),1ts<,(16)andgψϕ(ρ,s)=(logρ)ϕ+ψ1(log(ρ/s))ϕ+ψ1,1sρ<,(logρ)ϕ+ψ1,1ρs<,(17)with Φ ∈ {p, q}, ψ ∈ {αj, βi}, ρ ∈ {ξ, η}.

Proof

By Lemma 2.5, we have x(t)=1Γ(p)1t(logts)p1u(s)dss+(logt)p1[Γ(q)Ω1u(s)dssΓ(q)Ωi=1mλiΓ(q+αi)1η(logηs)q+αi1v(s)dss+Λ1Ω1v(s)dssΛ1Ωj=1nσjΓ(p+βj)1ξ(logξs)p+βj1u(s)dss]+1Γ(p)1(logt)p1u(s)dss(Γ(p)Γ(q)Λ1Λ2)ΩΓ(p)1(logt)p1u(s)dss=1gp(t,s)u(s)dss+Λ1Ω1(logt)p1v(s)dssΓ(q)Ωi=1mλi(logt)p1Γ(q+αi)1η(logηs)q+αj1v(s)dssΛ1Ωj=1nσj(logt)p1Γ(p+βj)1ξ(logξs)p+βj1u(s)dss+Λ1Λ2ΩΓ(p)1(logt)p1u(s)dss=1gp(t,s)u(s)dss+Γ(q)Ωi=1mλi(logt)p1Γ(q+αi)1(logη)q+αj1v(s)dssΓ(q)Ωi=1mλi(logt)p1Γ(q+αi)1η(logηs)q+αj1v(s)dssΛ1Ωj=1nσj(logt)p1Γ(p+βj)1ξ(logξs)p+βj1u(s)dss+Λ1Ωj=1nσj(logt)p1Γ(p+βj)1(logξ)p+βj1u(s)dss=1gp(t,s)u(s)dss+Λ1Ωj=1nσj(logt)p1Γ(p+βj)1gβjp(ξ,s)u(s)dss+Γ(q)Ωi=1mλi(logt)p1Γ(q+αi)1gαjq(η,s)v(s)dss=1G1(t,s)u(s)dss+1G2(t,s)v(s)dss, which yields that (14) is satisfied. In a similar way, we have y(t)=1Γ(q)1t(logts)q1v(s)dss+(logt)q1[Γ(p)Ω1v(s)dssΓ(p)Ωj=1nσjΓ(p+βj)1ξ(logξs)p+βj1u(s)dss+Λ2Ω1u(s)dssΛ2Ωi=1mλiΓ(q+αi)1η(logηs)q+αi1v(s)dss]+1Γ(q)1(logt)q1v(s)dss(Γ(p)Γ(q)Λ1Λ2)ΩΓ(q)1(logt)q1v(s)dss=1gq(t,s)v(s)dss+Λ2Ω1(logt)q1u(s)dssΓ(p)Ωj=1nσj(logt)q1Γ(p+βj)1ξ(logξs)p+βj1u(s)dssΛ2Ωi=1mλi(logt)q1Γ(q+αi)1η(logηs)q+αi1v(s)dss+Λ1Λ2ΩΓ(q)1(logt)q1v(s)dss=1gq(t,s)v(s)dss+Γ(p)Ωj=1nσj(logt)q1Γ(p+βj)1(logξ)p+βj1u(s)dssΓ(p)Ωj=1nσj(logt)q1Γ(p+βj)1ξ(logξs)p+βj1u(s)dssΛ2Ωi=1mλi(logt)q1Γ(q+αi)1η(logηs)q+αi1v(s)dss+Λ2Ωi=1mλi(logt)q1Γ(q+αi)1(logη)q+αi1v(s)dss=1gq(t,s)v(s)dss+Λ2Ωi=1mλi(logt)q1Γ(q+αi)1gαiq(η,s)v(s)dss+Γ(p)Ωj=1nσj(logt)q1Γ(p+βj)1ξgβjp(ξ,s)u(s)dss=1G3(t,s)v(s)dss+1G4(t,s)u(s)dss, which proves that (15) holds. This completes the proof.  □

Before establishing some properties of the Green’s functions, we set M1=1Γ(p)+Λ1Ωj=1nσj(logξ)p+Bj1Γ(p+βj),M2=Γ(q)Ωi=1mλi(logη)q+αi1Γ(q+αi),M3=1Γ(q)+Λ2Ωi=1mλi(logη)q+αi1Γ(q+αi),M4=Γ(p)Ωj=1nσj(logξ)p+Bj˙1Γ(p+βj).

Lemma 2.7

The Green’s functions Gi(t, s), i = 1, 2, 3, 4, satisfy the following properties:

  • (P1)Gi (t, s), i = 1, 2, 3, 4 are continuous for (t, s) ∈ [1, ∞) × [1,∞);

  • (P2)Gi(t, s) ≥ 0, i = 1, 2, 3, 4 for all (t, s) ∈ [1, ∞) × [1,∞);

  • (P3)G1(t,s)1+(logt)p1+(logt)q1M1,G2(t,s)1+(logt)p1+(logt)q1M2,G3(t,s)1+(logt)p1+(logt)q1M3,G4(t,s)1+(logt)p1+(logt)q1M4;

  • (P4)infξtξk1G1(t,s)1+(logt)p1+(logt)q1Λ1Ωj=1nσj(logξ)p1gβjp.(ξ,s)Γ(p+βj)(1+(logξ)p1+(logξ)q1),infξtξk1G2(t,s)1+(logt)p1+(logt)q1Γ(q)Ωj=1mλi(logη)p1gαiq(η,s)Γ(q+αi)(1+(logη)p1+(logη)q1),infξtξk1G3(t,s)1+(logt)p1+(logt)q1Λ2Ωj=1mλi(logη)q1gαiq(η,s)Γ(q+αi)(1+(logη)p1+(logη)q1),infξtξk1G4(t,s)1+(logt)p1+(logt)q1Γ(p)Ωj=1nσj(logξ)q1gβjp.(ξ,s)Γ(p+βj)(1+(logξ)p1+(logξ)q1),foranyκ1,κ2>1.

Proof

It is easy to prove that (P1) and (P2) hold.

To prove (P3), for (s, t) ∈ [1, ∞) × [1, ∞), we have G1(t,s)1+(logt)p1+(logt)q1=gp(t,s)1+(logt)p1+(logt)q1+Λ1Ωj=1nσj(logt)p1gβjp(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1)(logt)p1Γ(p)(1+(logt)p1+(logt)q1)+Λ1Ωj=1nσj(logt)p1gβjp(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1)1Γ(p)+Λ1Ωj=1nσjgβjp(ξ,s)Γ(p+βj).

In similar manner, we can prove that G2(t,s)1+(logt)p1+(logt)q1<_Γ(q)Ωi=1nλigαiq(η,s)Γ(q+αi),G3(t,s)1+(logt)p1+(logt)q1<_1Γ(q)+Λ2Ωi=1mλigαiq(η,s)Γ(q+αi),G4(t,s)1+(logt)p1+(logt)q1<_Γ(p)Ωj=1nσjgβjp(ξ,s)Γ(p+βj).

From (17), it follows that (P3) holds.

To prove (P4), from the positivity of gp(t, s) and gβjp(ξ,s), j = 1,..., n, we have for any κ1 > 1, minξtξκ1G1(t,s)1+(logt)p1+(logt)q1=minξtξκ1[gp(t,s)1+(logt)p1+(logt)q1+Λ1Ωj=1nσj(logt)p1gβjp.(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1)]minξtξκ1[Λ1Ωj=1nσj(logt)p1gβjp.(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1)]Λ1Ωj=1nσj(logξ)p1gβjp.(ξ,s)Γ(p+βj)(1+(logξ)p1+(logξ)q1).

In the same way, for any κ2 > 1, we have minηtηκ2G2(t,s)1+(logt)p1+(logt)q1Γ(q)Ωi=1mλi(logη)p1gαiq(η,s)Γ(q+αi)(1+(logη)p1+(logη)q1),minηtηκ2G3(t,s)1+(logt)p1+(logt)q1Λ2Ωi=1mλi(logη)q1gαiq(η,s)Γ(q+αi)(1+(logη)p1+(logη)q1),minηtηκ1G4(t,s)1+(logt)p1+(logt)q1Γ(p)Ωj=1nσj(logξ)q1gβjp.(ξ,s)Γ(p+βj)(1+(logξ)p1+(logξ)q1).

Therefore, (P4) is proved.  □

3 Main results

Define the set E={(x,y)C([1,),R)×C([1,),R):supt[1,)|x(t)|+|y(t)|1+(logt)p1+(logt)q1<}, and the norm (x,y)=supt[1,)|x(t)|+|y(t)|1+(logt)p1+(logt)q1.

It is clear that (E, || • ||) is a Banach space.

Lemma 3.1

Let U ⊂ E be a bounded set. If the following conditions hold:

  1. for any (u,v)(t)U,u(t)+v(t)1+(logt)p1+(logt)q1is equicontinuous on any compact interval of [1, ∞);

  2. for any ε > 0, there exists a constant T = T(ε) > 0 such thatu(t1)+v(t1)1+(logt1)p1+(logt1)q1u(t2)+v(t2)1+(logt2)p1+(logt2)q1<ϵ(18)for any t1,t2 ≥ T and (u,v) ∈ U,

    then U is relatively compact in E.

The proof is similar to that of Lemma 2.8 in [31], and is omitted.

Now, we define the positive cone PE by P={(x,y)E:x(t)0,y(t)0on[1,)}, and the operator 𝒯 : PE by 𝒯(x, y)(t) = (A(x, y)(t), B(x, y)(t)) for all t ∈ [1, ∞), where the operators A : PE and B : PE are defined by A(x,y)(t):=1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss,B(x,y)(t):=1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss.(19)

Throughout this paper we assume that the following conditions hold:

  • (H1)The functions f, gC([0,∞) × [0, ∞), [0, ∞)), f(x, y), g(x, y) ≠ 0 on any subinterval of (0, ∞) × (0, ∞) and f, g are bounded on [0, ∞) × [0, ∞);

  • (H2)a, b : [1, ∞) → [1, ∞) are not identical zero on any closed subinterval of [1, ∞) and 0<1a(s)dss<,0<1b(s)dss<.

Next, we are going to prove that the operator 𝒯 is completely continuous.

Lemma 3.2

Let (H1) and (H2) hold. Then 𝒯 : P → P is completely continuous.

Proof

Firstly, we will show that 𝒯 is uniformly bounded on P. Let Ω* = {(x, y) ∈ P : ||(x, y)|| ≤ r} ≤ P be a bounded set. From (H1), there exist two positive constants L1 and L2 such that |f(x(t),y(t))|L1,|g(x(t),y(t))|L2,(x,y)Ω.

Then, from (H2), for any (x, y) ∈ Ω*, we have T(x,y)=supt[1,)11+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]M11a(s)f(x(s),y(s))dss+M21b(s)g(x(s),y(s))dss+M31b(s)g(x(s),y(s))dss+M41a(s)f(x(s),y(s))dss(M1+M4)L11a(s)dss+(M2+M3)L21b(s)dss<.

This means that the operator 𝒯 is uniformly bounded.

Secondly, we will show that 𝒯 is equicontinuous on any compact interval of [1, ∞). For any S > 1, t1, t2 ∈ [1, S], and (x, y) ∈ Ω*, without loss of generality, we assume that t1 < t2. Then, we have |A(x,y)(t2)1+(logt2)p1+(logt2)q1A(x,y)(t1)1+(logt1)p1+(logt1)q1|=|1G1(t2,s)1+(logt2)p1+(logt2)q1a(s)f(x(s),y(s))dss+1G2(t2,s)1+(logt2)p1+(logt2)q1b(s)g(x(s),y(s))dss1G1(t1,s)1+(logt1)p1+(logt1)q1a(s)f(x(s),y(s))dss1G2(t1,s)1+(logt1)p1+(logt1)q1b(s)g(x(s),y(s))dss||1(G1(t2,s)1+(logt2)p1+(logt2)q1G1(t1,s)1+(logt1)p1+(logt1)q1)a(s)f(x(s),y(s))dss|+|1(G2(t2,s)1+(logt2)p1+(logt2)q1G2(t1,s)1+(logt1)p1+(logt1)q1)b(s)g(x(s),y(s))dss|.

Since |G1(t2,s)1+(logt2)p1+(logt2)q1G1(t1,s)1+(logt1)p1+(logt1)q1||gp(t2,s)gp(t1,s)1+(logt2)p1+(logt2)q1+|(logt2)p1(logt1)p1+(logt2)ql(logt1)ql|gp(t1,s)(1+(logt2)p1+(logt2)q1)(1+(logt1)p1+(logt1)q1)+|(logt2)p1(logt1)p1+(logt2)p1(logt1)q1(logt1)p1(logt2)q1|(1+(logt2)p1+(logt2)q1)(1+(logt1)p1+(logt1)q1)Λ1Ωj=1nσjgβjp(ξ,s)Γ(p+βj), and 1|gp(t2,s)gp(tl,s)|1+(logt2)p1+(logt2)q1a(s)f(x(s),y(s))dss1t1|(logt2)p1(logt1)p1|+|(log(t2/s))p1(log(t1/s))p1|Γ)(1+(logt2)p1+(logt2)q1)a(s)f(x(s),y(s))dss+t1t2|(logt2)p1(logt1)p1|+|(log(t2/s))p1|Γ(p)(1+(logt2)p1+(logt2)q1)a(s)f(x(s),y(s))dss+t2|(logt2)p1(logt1)p1|Γ(p)(1+(logt2)p1+(logt2)q1)a(s)f(x(s),y(s))dss, we have that 1(G1(t2,s)1+(logt2)p1+(logt2)q1G1(t1,s)1+(logt1)p1+(logt1)q1)a(s)f(x(s),y(s))dss0, uniformly as t1t2. In addition, we can find that G2(t2,s)1+(logt2)p1+(logt2)q1G2(t1,s)1+(logt1)p1+(logt1)q1(|(logt2)p1(logt1)p1|+|(logt2)p1(logt1)q1(logt1)p1(logt2)q1|)(1+(logt2)p1+(logt2)q1)(1+(logt1)p1+(logt1)q1)×Γ(q)Ωi=1mλigαiq(η,s)Γ(q+αi), which leads to 1(G2(t2,s)1+(logt2)p1+(logt2)q1G2(t1,s)1+(logt1)p1+(logt1)q1)b(s)g(x(s),y(s))dss0, uniformly as t1t2. Hence A(x,y)(t2)1+(logt2)p1+(logt2)q1A(x,y)(t1)1+(logt1)p1+(logt1)q10,uniformlyast1,t2.

In the similar way, we can prove that B(x,y)(t2)1+(logt2)p1+(logt2)q1B(x,y)(t1)1+(logt1)p1+(logt1)q10, as t1t2 independently of (x, y) ∈ Ω*. Therefore 𝒯Ω* is equicontinuous on [1, ∞).

Thirdly, we will show that the operator 𝒯 is equiconvergent at ∞. From the first step, for any (x, y) ∈ Ω*, we have 1a(s)f(x(s),y(s))dssL11a(s)dss<, and 1b(s)g(x(s),y(s))dssL21b(s)dss<.

Then limt|A(x,y)(t)1+(logt)p1+(logt)q1|=limt|1G1(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dss+1G2(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss|limt1((logt)p1Γ(p)(1+(logt)p1+(logt)q1)+Λ1Ωj=1nσj(logt)p1gβjp.(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1))a(s)f(x(s),y(s))dss+limt1(Γ(q)Ωi=1mλi(logt)p1gαiq(η,s)Γ(q+αi)(1+(logt)p1+(logt)q1))b(s)g(x(s),y(s))dssM1L11a(s)dss+M2L21b(s)dss<, and limt|B(x,y)(t)1+(logt)p1+(logt)q1|=limt|1G3(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss+1G4(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dss|limt1((logt)q1Γ(q)(1+(logt)p1+(logt)q1)+Λ2Ωi=1mλi(logt)q1gαiq(η,s)Γ(q+αi)(1+(logt)p1+(logt)q1))b(s)g(x(s),y(s))dss+limt1(Γ(p)Ωj=1nσj(logt)q1gβjp.(ξ,s)Γ(p+βj)(1+(logt)p1+(logt)q1))a(s)f(x(s),y(s))dssM3L21b(s)dss+M4L11a(s)dss<, which imply limtT(x,y)(t)1+(logt)p1+(logt)q1<. Hence 𝒯Ω* is equiconvergent at infinity.

Finally, we will prove that the operator 𝒯 is continuous. Let (xn, yn)→ ∞ as n → ∞ in P. By applying the Lebesgue dominated convergence and the continuity of f and g guarantee that 1a(s)f(xn(s),yn(s))dss1a(s)f(x(s),y(s))dssasn, and 1b(s)g(xn(s),yn(s))dss1b(s)g(x(s),y(s))dssasn. Therefore, we get T(xn,yn)T(x,y)=(A(xn,yn)(t)A(x,y)(t),B(xn,yn)(t)B(x,y)(t))=supt[1,)|A(xn,yn)(t)A(x,y)(t)|+|B(xnyn)(t)B(x,y)(t)|1+(logt)p1+(1ogt)q1M11a(s)|f(xn(s),yn(s))f(x(s),y(s))|dss+M21b(s)|g(xn(s),yn(s))g(x(s),y(s))|dss+M31b(s)|g(xn(s),yn(s))g(x(s),y(s))|dss+M41a(s)|f(xn(s),yn(s))f(x(s),y(s))|dss0asn. Thus, the operator 𝒯 is continuous.

By applying Lemma 3.1, we deduce that the operator 𝒯 : PP is completely continuous. This completes the proof.  □

To prove our main results, we set the following constants m1=Λ1Ωj=1nσj(logξ)2p+Bj2Γ(p+βj)(1+(logξ)p1+(logξ)q1),m2=Γ(q)Ωi=1mλi(logη)p+q+αi2Γ(q+αi)(1+(logη)p1+(logη)q1),m3=Λ2Ωi=1mλi(logη)2q+αi2Γ(q+αi)(1+(logη)p1+(logη)q1),m4=Γ(p)Ωj=1nσj(logξ)p+q+Bj2Γ(p+βj)(1+(logξ)p1+(logξ)q1),n1=1a(s)dss,n2=1b(s)dss,n3=ξk1ξa(s)dss,n4=ηk2ηb(s)dss,Λ3=n3(m1+m4),Λ4=n4(m2+m3),Λ5=n1(M1+M4),Λ6=n2(M2+M3).

3.1 Existence result via Guo-Krasnoselskii’s fixed point theorem

In this subsection, the existence theorems of at least one positive solution will be established using the Guo-Krasnoselskii fixed point theorem.

Theorem 3.3

(Guo-Krasnoselskii fixed point theorem) [39] Let E be a Banach space, and let 𝒫E be a cone. Assume that Φ1, Φ2 are bounded open subsets of E with 0 ∈ ϕ1, ̄ϕ1 ⊂ Φ2, and let Q : 𝒫 ∩ (̄Φ2 \ Φ1) → 𝒫 be a completely continuous operator such that:

  1. ||Qx||≥||x||, x𝒫 ∩ ∂ Φ1, and ||Qx||≤||x||, x𝒫Φ2; or

  2. ||Qx||≤||x||, x𝒫 ∩ ∂ Φ1, and ||Qx||≥||x||, x𝒫Φ2.

Then Q has a fixed point in 𝒫 ∩ (̄Φ21).

Theorem 3.4

let f, g : + × ++ be continuous functions. Suppose that there exist positive constants ρ1<ρ2,andθ1(Λ31,),θ2(Λ41,),θ3(0,Λ51)andθ4(0,Λ61). In addition, assume that the following conditions hold: (H3)f(x,y)θ1ρ12for(x,y)[0,ρ1]×[0,ρ1]andg(x,y)θ2ρ12for(x,y)[0,ρ1]×[0,ρ1];(H4)f(x,y)θ3ρ22for(x,y)[0,ρ2]×[0,ρ2]andg(x,y)θ4ρ22(x,y)[0,ρ2]×[0,ρ2].Then, the problem (4) has at least one positive solution (x, y) such thatρ1<(x,y)<ρ2.

Proof

It follows from Lemma 3.2 that the operator 𝒯 : 𝒫𝒫 is completely continuous. Define Φ1 = {(x, y) ∈ E : ||(x, y)|| < ρ1}. Hence, for any (x, y) ∈ 𝒫Φ1, we have 0 ≤ x(t) ≤ ρ1 and 0 ≤ y(t) ≤ ρ1 for all t ∈ [1, ∞). From assumption (H3) and Lemma 2.7, we get T(x,y)=supt[1,)|A(x,y)(t)|+|B(x,y)(t)|1+(logt)p1+(logt)q1=supt[1,)11+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]1inft[ξ,k1ξ]G1(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dss+1inft[η,k2η]G2(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1inft[η,k2η]G3(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1inft[ξ,k1ξ]G4(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dssΛ1Ωj=1nσj(logξ)p1Γ(p+βj)(1+(logξ)p1+(logξ)q1)ξk1ξgβjp(ξ,s)a(s)f(x(s),y(s))dss+Γ(q)Ωi=1mλi(logη)p1Γ(q+αi)(1+(logη)p1+(logη)q1)ηk2ηgαiq(η,s)b(s)g(x(s),y(s))dss+Λ2Ωi=1λi(logη)q1Γ(q+αi)(1+(logη)p1+(logη)q1)mηk2ηgαiq(η,s)b(s)g(x(s),y(s))dss+rτ(p)Ωj=1nσj(logξ)q1Γ(p+βj)(1+(logξ)p1+(logξ)q1)ξk1ξgβjp(ξ,s)a(s)f(x(s),y(s))dss12m1ρ1θ1ξk1ξa(s)dss+12m2ρ1θ2ηk2ηb(s)dss+12m3ρ1θ1ηk2ηb(s)dss+12m4ρ1θ2ξk1ξa(s)dss=12Λ3θ1ρ1+12Λ4θ2ρ1ρ1, which implies that ||𝒯(x,y)|| ≥ ||(x,y)|| for (x,y) ∈ 𝒫 ∩ ∂Φ1.

Next, we define Φ2 = {(x,y) ∈ E : ||(x,y)|| < ρ2}. Hence, for any (x,y) ∈ 𝒫 ∩∂Φ2, we have 0 ≤ x(t) ≤ ρ2 and 0 ≤ y(t) ≤ ρ2 for all t ∈ [1,∞). Using the condition (H4), we obtain T(x,y)=supt[1,)11+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]12M1θ3ρ21a(s)dss+12M2θ4ρ21b(s)dss+12M3θ4ρ21b(s)dss+12M4θ3ρ21a(s)dss=12Λ5θ3ρ2+12Λ6θ4ρ2ρ2, which leads to ||𝒯(x,y)|| ≤ ||(x,y)|| for (x,y) ∈ 𝒫 ∩ ∂Φ2.

Therefore, by the first part of Theorem 3.3, we deduce that the operator 𝒯 has a fixed point in 𝒫 ∩ (̄Φ21) which is a positive solution of problem (4). Thus the problem (4) has at least one positive solution (x, y) such that ρ1<(x,y)<ρ2. This completes the proof. □

Similarly to the previous theorem, we can prove the following result.

Theorem 3.5

Let f, g : ℝ+ × ℝ+ → ℝ+ be continuous functions. Assume that there exist positive constants ρ1<ρ2,andθ1(Λ31,),θ2(Λ41,),θ3(0,Λ51)andθ4(0,Λ61). In addition, assume that the following conditions hold:

(H5)f(x,y)θ3ρ12for(x,y)[0,ρ1]×[0,ρ1]andg(x,y)θ4ρ12for(x,y)[0,ρ1]×[0,ρ1];

(H6)f(x,y)θ1ρ22for(x,y)[0,ρ2]×[0,ρ2]andg(x,y)θ2ρ22for(x,y)[0,ρ2]×[0,ρ2].

Then, the problem (4) has at least one positive solution (x, y) such thatρ1<(x,y)<ρ2.

3.2 Existence result via Leggett-Williams fixed point theorem

In this subsection, the existence of at least three positive solutions will be proved using the Leggett-Williams fixed point theorem.

Definition 3.6

A continuous mapping ω : 𝒫 → [0,∞) is said to be a nonnegative continuous concave functional on the cone 𝒫 of a real Banach space E provided thatω(λx+(1λ)y)λω(x)+(1λ)ω(y)for all x,y𝒫 and λ ∈ [0,1].

Let a, b,d > 0 be given constants and define Pd={(x,y)P:(x,y)<d},P¯d={(x,y)P:(x,y)d}andP(ω,a,b)={(x,y)P:ω((x,y))a,(x,y)b}.

Theorem 3.7

(Leggett-Williams fixed point theorem) [40] Let 𝒫 be a cone in the real Banach space E and c>0 be a constant. Assume that there exists a concave nonnegative continuous functional ω on 𝒫 with ω (x) ≤ ||x|| for all x ∈ ̄𝒫c. Let Q : ̄𝒫c → ̄𝒫c be a completely continuous operator. Suppose that there exist constants 0 < a < b < dc such that the following conditions hold:

  1. {xP(ω,b,d):ω(x)>b}andω(Qx)>bforxP(ω,b,d);

  2. Qx<aforxa;

  3. ω(Qx)>bforxP(ω,b,c)withQx>d.

Then Q has at least three fixed points x1, x2 and x3 in ̄𝒫c. In addition, ||x1|| < a, ω(x2) > b, ||x3|| > a with ω(x3) < b.

Theorem 3.8

Let f, g : ℝ+ × ℝ+ → ℝ+ be continuous functions. Suppose that there exist two constants k1, k2 > 1 such that [ξ, k1 ξ] ∩ [η, k2η] ≠ θ. In addition, assume that there exist positive constants a < b < c satisfying

(H7)f(x,y)<a2Λ5andg(x,y)>a2Λ6for(x,y)[0,a]×[0,a];

(H8)f(x,y)<b2Λ3andg(x,y)>b2Λ4for(x,y)[b,c]×[b,c];

(H9)f(x,y)<c2Λ5andg(x,y)>b2Λ6for(x,y)[0,c]×[0,c].

Then, the problem (4) has at least three positive solutions (x1, y1), (x2, y2) and (x3, y3) such that (x1,y1)<a,infτ1tτ2(x2,y2)(t)>band(x3,y3)>awithinfτ1tτ2(x3,y3)(t)<b,whereτ1=max{ξ,η},τ2=min{k1ξ,k2η}.

Proof

At first, we will show that the operator 𝒯 : ̄𝒫c → ̄𝒫c. For any (x, y) ∈ ̄𝒫c, we have ||(x, y)|| ≤ c. Using the condition (H9) and Lemma 2.7, we obtain T(x,y)=supt[1,)11+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]1supt[1,)G1(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dss+1supt[1,)G2(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss+1supt[1,)G3(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss+1supt[1,)G4(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dssc2Λ5M1n1+c2Λ6M2n2+c2Λ6M3n2+c2Λ5M4n1=c, which yields 𝒯 : ̄𝒫c → ̄𝒫c. Secondly, we let (x, y) ∈ ̄Pa. By applying condition (H1 ), we have T(x,y)1supt[1,)G1(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dss+1supt[1,)G2(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss+1supt[1,)G3(t,s)1+(logt)p1+(logt)q1b(s)g(x(s),y(s))dss+1supt[1,)G4(t,s)1+(logt)p1+(logt)q1a(s)f(x(s),y(s))dss<a2Λ5M1n1+a2Λ6M2n2+a2Λ6M3n2+a2Λ5M4n1=a. This means that the condition (i i) of Theorem 3.7 is satisfied.

Thirdly, we define τ1=max{ξ,η},τ2=min{k1ξ,k2η} and also a concave nonnegative continuous functional ω on E by ω((x,y))=infτ1tτ2|x(t)|+|y(t)|1+(logt)p1+(logt)q1. By choosing (x,y)(t)=((b+c)2(1+(logt)p1+(logt)q1),(b+c)2(1+(logt)p1+(logt)q1)), we deduce that (x, y)(t) ∈ ̄𝒫(ω, b, c) and ω((x, y)) > b. Hence it follows that {(x, y) ∈ ̄𝒫(ω, b, c) : ω((x, y)) > b} ≠ Ø. Therefore, if (x, y) ∈ ̄𝒫(ω, b, c), then we have bx(t) ≤ c and by(t) ≤ c for t ∈ [τ1, τ2]. Using condition (H8) and Lemma 2.7, we obtain ω(T(x,y)(t))=infτ1tτ2|A(x,y)(t)|+|B(x,y)(t)|1+(logt)p1+(logt)q1=infτ1tτ211+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]1infτ1tτ2G1(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G2(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G3(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G4(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dss>b2Λ3m1n3+b2Λ4m2n4+b2Λ4m3n4+b2Λ3m4n3=b. Therefore, ω(𝒯(x,y)) > b for all (x,y) ∈ P(ω,b,c). This means that the condition (i) of Theorem 3.7 is fulfilled.

Finally, we assume that (x,y)(t) ∈ 𝒫(ω,b,c) with ||𝒯 (x,y)|| > d, where b < dc. Then, we have bx(t) ≤ c and by(t) ≤ c for t ∈ [τ1, τ2]. From condition (H8) and Lemma 2.7, we get ω(T(x,y)(t))=infτ1tτ211+(logt)p1+(logt)q1[1G1(t,s)a(s)f(x(s),y(s))dss+1G2(t,s)b(s)g(x(s),y(s))dss+1G3(t,s)b(s)g(x(s),y(s))dss+1G4(t,s)a(s)f(x(s),y(s))dss]1infτ1tτ2G1(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G2(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G3(t,s)b(s)g(x(s),y(s))1+(logt)p1+(logt)q1dss+1infτ1tτ2G4(t,s)a(s)f(x(s),y(s))1+(logt)p1+(logt)q1dss>b2Λ3n3(m1+m4)+b2Λ4n4(m2+m3)=b. It follows that the condition (iii) of Theorem 3.7 is satisfied. Hence, by applying Theorem 3.7, we deduce that the problem (4) has at least three positive solutions (x1, y1), (x2, y2) and (x3, y3) such that ||(x1, y1|| < a, infτ1tτ2(x2,y2)(t)>band(x3,y3)>awithinfτ1tτ2(x3,y3)(t)<b. This completes the proof.  □

4 Examples

In this section, we present two examples to illustrate our results.

Example 4.1

Consider the following Hadamard fractional differential system subject to boundary conditions on an unbounded domainD3/2x(t)+erf(x(t),y(t))=0,t(1,),D5/3y(t)+t2g(x(t),y(t))=0,t(1,),x(1)=0,D1/2x()=23I1/2y(74)+π5I3/2y(74),y(1)=0,D2/3y()=25I1/3x(83)+17I2/3x(83)+2eI4/3x(83),(20)wheref(x,y)=x1+y(14x)+y1+x(14y)+7;0x,y1/4,6+e116xy+sin2((14x)(14y));1/4x,y<,andg(x,y)={x21+x2(14y)+y41+y4(14x)+2;0x,y1/4,2+exysin4((x14)(y14));1/4x,y<. Here p=3/2,q=5/3,a(t)=et,b(t)=t2,m=2,η=7/4,λ1=2/3,α1=1/2,λ2=π/5,α2=3/2,n=3,ξ=8/3,σ1=2/5,β1=1/3,σ2=1/7,β2=2/3,σ3=2/eandβ3=4/3. We find that Λ1 = 0.3512388401 and Λ2 = 0.9782224108 which leads to Ω = 0.4564474804 > 0. In addition, we can compute that M1 = 1.977763776, M2 = 0.7695054857, M3 = 1.941574800, M4 = 2.143121504, m1 = 0.2825156640, m2 = 0.2371688476, m3 = 0.2332983559, m4 = 0.7105323246, n1 = 0.2193839344 and n2 = 0.5000000000. By choosing k1 = 2 and k2 = 3, we also obtain n3 = 0.01925300492, n4 = 0.1451247166, Λ3 = 0.01911915781, Λ4 = 0.06827641958, Λ5 = 0.9040560259 and Λ6 = 1.355540143. Observe that the functions f, g, a and b satisfy the conditions .(H1)-(H2).

Choosing ρ1 = 1/4, ρ2 = 100, θ1 = 54 ∈ (Λ-13, ∞) = (52.30355908, ∞), θ2 = 16 ∈ (Λ-14, ∞) = (14.64634505, ∞) θ3 = 1 ∈ (0, Λ-15) = (0, 1.106126137) and θ4 = (0.5 ∈ (0, Λ-16) = (0, 0.7377133058), we obtain f(x,y)7θ1ρ12andg(x,y)2θ2ρ12, for 0 ≤ x,y ≤ 1/4. Also we have f(x,y)50θ3ρ22andg(x,y)25θ4ρ22, for 1/4 ≤ x, y < ∞.

Hence the conditions (H3)-(H4) hold. By Theorem 3.4, we conclude that the problem (20) has at least one positive solution (x,y) such that 1/4 < ||(x,y)|| < 100.

Example 4.2

Consider the following Hadamard fractional differential system subject to boundary conditions on an unbounded domainD7/4x(t)+t3/4f(x(t),y(t))=0,t(1,),D9/5y(t)+e2tg(x(t),y(t))=0,t(1,),x(1)=0,D3/4x()=π13I1/4y(95)+712I1/2y(95)+215I3/4y(95)y(1)=0,D4/5y()=316I1/5x(73)+25I2/5x(73)+13e2I3/5x(73)+38πI4/5X(73),(21) wheref(x,y)=x5(1+x)(35x)+y5(1+y)(35y)+3100;0x,y3/5,3100e|xy|+3(x35)+2(y35);3/5x,y7/5,403100+e((7/5)y)sin2(75x);7/5x,y<, and g(x,y)=xy(35x)(35y)+2;0x,y3/5,2e|x2y2|+230(x35)+220(y35);3/5x,y7/5,362+e2((7/5)x)sin4(75y);7/5x,y<. Here p=7/4,q=9/5,a(t)=t3/4,b(t)=e2t,m=3,η=9/5,λ1=π/13,α1=1/4,λ2=7/12,α2=1/2,λ3=2/15,α2=3/4,n=4,ξ=7/3,σ1=3/16,β1=1/5,σ2=2/5,β2=2/5,σ3=1/e2 and β3 = 3/5, σ4 = 3/8ᅲ and β4 = 4/5. We find that Λ1 = 0.3324581038 and Λ2 = 0.8725814055 which leads to Ω = 0.5659031627 > 0. In addition, we can compute that M1 = 1.645835954, M2 = 0.5874823216, M3 = 1.624063246, M4 = 1.541927070, m1 = 0.1785392532, m2 = 0.1696242483, m3 = 0.1547484152, m4 = 0.4894898091, n1 = 1.333333333 and n2 = 0.04890051071. By choosing k1 = 4 and k2 = 5, it is easy to see that [ξ, k1ξ] ∩ [η, k2η] ≠θ. Then we also obtain n3 = 0.4565504861, n4 = 0.006160413528, Λ3 =0.3049889931, Λ4 = 0.001998269744, Λ5 = 4.250350698 and Λ6 = 0.1081457077. From above information, the conditions (H1)-(H2) are fulfilled.

Choosing a = 3/5, b = 7/5, c = 80, we get f(x,y)0.06600000000andg(x,y)2.008100000000, which yields for 0 ≤ x, y ≤ 3/5, f(x,y)<0.07058241103=a2Λ5andg(t,x,y)<2.774035201=a2Λ6. In addition, we obtain f(x,y)4.030000000andg(x,y)362.000000, which leads to f(x,y)>2.295164796=b2Λ3andg(x,y)>350.3030570=b2Λ4, for 7/5 ≤ x, y ≤ 80. Also we have for 0 ≤ x, y ≤ 80. f(x,y)9.410988138=c2Λ5andg(x,y)369.8713601=c2Λ6. It is easy to see that τ1=max{ξ,η}=7/3,τ2=min{k1ξ,k2η}=9.

Therefore, the conditions (H7)-(H9) of Theorem 3.8 hold. Applying Theorem 3.8, we deduce that the problem (21) has at least three positive solutions (x1, y1), (x2, y2) and (x3, y3) such that ||(x1, y1|| < 3/5, inf7/3t9(x2,y2)(t)>7/5and(x3,y3)>3/5withinf7/3t9(x3,y3)(t)<7/5.

Acknowledgement

This research was funded by King Mongkut’s University of Technology North Bangkok. Contract no. KMUTNB-GEN-59-18.

References

  • [1]

    Miller K.S., Ross B., An Introduction to the Fractional Calculus and Differential Equations, John Wiley, New York, 1993. Google Scholar

  • [2]

    Samko S.G., Kilbas A.A., Marichev O.I., Fractional Integrals and Derivatives, Theory and Applications, Gordon and Breach, Yverdon, 1993.Google Scholar

  • [3]

    Podlubny I., Fractional Differential Equations, Academic Press, San Diego, 1999.Google Scholar

  • [4]

    Kilbas A.A., Srivastava H.M., Trujillo J.J., Theory and Applications of Fractional Differential Equations, North-Holland Mathematics Studies, 204. Elsevier Science B.V., Amsterdam, 2006. Google Scholar

  • [5]

    Lakshmikantham V., Leela S., Devi J.V., Theory of Fractional Dynamic Systems, Cambridge Academic Publishers, Cambridge, 2009.Google Scholar

  • [6]

    Baleanu D., Diethelm K., Scalas E., Trujillo J.J., Fractional calculus models and numerical methods. Series on Complexity, Nonlinearity and Chaos, World Scientific, Boston, 2012.Google Scholar

  • [7]

    Diethelm K., The analysis of fractional differential equations. An application-oriented exposition using differential operators of Caputo type, Lecture Notes in Mathematics 2004, Springer, 2010. Google Scholar

  • [8]

    Guezane-Lakoud A., Khaldi R., Solvability of a three-point fractional nonlinear boundary value problem, Differ. Equ. Dyn. Syst.20 (2012), 395-403. CrossrefGoogle Scholar

  • [9]

    Kaufmann E., Existence and nonexistence of positive solutions for a nonlinear fractional boundary value problem, Discrete Contin. Dyn. Syst. (2009), Suppl., 416-423.Google Scholar

  • [10]

    Wang J., Xiang H.. Liu Z., Positive solution to nonzero boundary values problem for a coupled system of nonlinear fractional differential equations, Int. j. Differ. Equ. 2010, Article ID 186928, 12 p.Google Scholar

  • [11]

    Bai Z., Lü H.. Positive solutions for boundary value problem of nonlinear fractional differential equation, j. Math. Anal. Appl. 311 (2005), 495-505.CrossrefGoogle Scholar

  • [12]

    Bai Z., On positive solutions of a nonlocal fractional boundary value problem, Nonlinear Anal. 72 (2010), 916-924.Web of ScienceGoogle Scholar

  • [13]

    Sudsutad W., Tariboon j., Boundary value problems for fractional differential equations with three-point fractional integral boundary conditions, Adv. Difference Equ. (2012) 2012:93Google Scholar

  • [14]

    Ntouyas S.K., Existence results for nonlocal boundary value problems for fractional differential equations and inclusions with fractional integral boundary conditions, Disc. Math. Diff. Incl. Contr. Optim. 33 (2013), 17-39. CrossrefGoogle Scholar

  • [15]

    Yang W., Positive solutions for nonlinear Caputo fractional differential equations with integral boundary conditions, j. Appl. Math. Comput. DOI 10.1007/s12190-013-0679-8. Google Scholar

  • [16]

    Bai Y.R., Hadamard fractional calculus for interval-valued functions, j. Comput. Complex. Appl. 3 (1) (2017) 23-43.Google Scholar

  • [17]

    He W., Adomian decomposition method for fractional differential equations of Caputo Hadamard type, j. Comput. Complex. Appl. 2(4) (2016) 160-162.Google Scholar

  • [18]

    Ahmad B., Ntouyas S.K., Assolani A., Caputo type fractional differential equations with nonlocal Riemann-Liouville integral boundary conditions, j. Appl. Math. Comput. 41 (2013), 339-350. CrossrefGoogle Scholar

  • [19]

    Agarwal R.P., O’Regan D., Infinite Interval Problems for Differential, Difference and Integral Equations, Kluwer Academic Publisher, Dordrecht, 2001. Google Scholar

  • [20]

    O’Regan D., Theory of Singular Boundary Value Problems, World Scientific, River Edge, New Jersey, USA, 1994. Google Scholar

  • [21]

    Liu Y., Existence and unboundedness of positive solutions for singular boundary value problems on half-line, Appl. Math.Comput. 144(2003)543-556. Google Scholar

  • [22]

    Liang S., Zhang j., Existence of multiple positive solutions for m-point fractional boundary value problems on an infinite interval, Math. Comput. Modelling 54 (2011), 1334-1346.Web of ScienceCrossrefGoogle Scholar

  • [23]

    Liu Y.S., Boundary value problem for second order differential equations on unbounded domain, Acta Anal. Funct. Appl. 4 (2002), 211-216.Google Scholar

  • [24]

    Zhang L.. Ahmad B., Wang G., Agarwal R.P., Al-Yami M., Shammakh W., Nonlocal integrodifferential boundary value problem for nonlinear fractional differential equations on an unbounded domain, Abstr. Appl. Anal. Volume 2013, Article ID 813903, 5 pages.Web of ScienceGoogle Scholar

  • [25]

    Zhao X., Ge W., Unbounded solutions for a fractional boundary value problems on the infinite interval, Acta Appl. Math. 109 (2010), 495-505. Web of ScienceCrossrefGoogle Scholar

  • [26]

    Ntouyas S.K., Obaid M., A coupled system of fractional differential equations with nonlocal integral boundary conditions, Adv. Difference Equ (2012), 2012:130. Google Scholar

  • [27]

    Ahmad B., Ntouyas S.K., A fully Hadamard-type integral boundary value problem of a coupled system of fractional differential equations, Fract. Calc. Appl. Anal. 17(2014), 348-360.Web of ScienceGoogle Scholar

  • [28]

    Ahmad B., Ntouyas S.K., Existence results for a coupled system of Caputo type sequential fractional differential equations with nonlocal integral boundary conditions, Appl. Math. Comput. 266 (2015), 615-622.Web of ScienceGoogle Scholar

  • [29]

    Sudsutad W., Ntouyas S.K., Tariboon j., Systems of fractional Langevin equation via Riemann-Liouville and Hadamard types and their fractional integral conditions, Adv. Difference Equ. (2015) 2015:235.Google Scholar

  • [30]

    Tariboon j., Ntouyas S.K., Sudsutad W., Coupled systems of Riemann-Liouville fractional differential equations with Hadamard fractional integral boundary conditions, j. Nonlinear Sci. Appl. 9 (2016), 295-308. CrossrefGoogle Scholar

  • [31]

    Thiramanus P., Ntouyas S.K., Tariboon j., Positive solutions for Hadamard fractional differential equations on infinite domain, Adv. Difference Equ. (2016) 2016:83. Web of ScienceGoogle Scholar

  • [32]

    Agarwal R.P., O’Regan D., Stanek S., Positive solutions for mixed problems of singular fractional differential equations, Math. Nachr. 285(2012), 27-41. CrossrefWeb of ScienceGoogle Scholar

  • [33]

    Goodrich C.S., Existence of a positive solution to systems of differential equations of fractional order, Comput. Math. Appl. 62 (2011) 1251-1268.CrossrefWeb of ScienceGoogle Scholar

  • [34]

    Wang Y., Positive solutions for a system of fractional integral boundary value problem, Bound. Value Prob. vol. 2013, article 256, 14 pages.Google Scholar

  • [35]

    Wang Y., Liu L. Y. Wu, Positive solutions for a class of higher-order singular semipositone fractional differential systems with coupled integral boundary conditions and parameters, Adv. Difference Equ. (2014) 2014:268.Google Scholar

  • [36]

    Henderson J., Luca R., Positive solutions for a system of fractional differential equations with coupled integral boundary conditions, Appl. Math. Comput. 249(2014), 182-197.CrossrefGoogle Scholar

  • [37]

    Henderson J., Luca R., Tudorache A., On a system of fractional differential equations with coupled integral boundary conditions, Fract. Calc. Appl. Anal. 18 (2015) 361-386. Web of ScienceGoogle Scholar

  • [38]

    Neamprem K., Muensawat T. Ntouyas S.K., j. Tariboon, Positive solutions for fractional differential systems with nonlocal Riemann-Liouville fractional integral boundary conditions, Positivity, DOI 10.1007/s11117-016-0433-1. Web of ScienceGoogle Scholar

  • [39]

    Guo D., Lakshmikanthan V., Nonlinear Problems in Abstract Cones, Academic Press, New York (1988) Google Scholar

  • [40]

    Leggett R.W., Williams L.R.. Multiple positive fixed points of nonlinear operators on ordered Banach spaces, Indiana Univ. Math. j. 28(1979), 673-688. CrossrefGoogle Scholar

About the article

Received: 2016-10-10

Accepted: 2017-03-21

Published Online: 2017-05-25


Citation Information: Open Mathematics, Volume 15, Issue 1, Pages 645–666, ISSN (Online) 2391-5455, DOI: https://doi.org/10.1515/math-2017-0057.

Export Citation

© 2017 Tariboon et al.. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. BY-NC-ND 3.0

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

[1]
Jinxiu Mao, Zengqin Zhao, and Chenguang Wang
Journal of Function Spaces, 2019, Volume 2019, Page 1

Comments (0)

Please log in or register to comment.
Log in