We prove the existence and uniqueness of solutions for two classes of infinite delay nonlinear fractional order differential equations involving Riemann-Liouville fractional derivatives. The analysis is based on the alternative of the Leray-Schauder fixed-point theorem, the Banach fixed-point theorem, and the Arzela-Ascoli theorem in Ω={y:(−∞,b]→ℝ:y|(−∞,0]∈ℬ} such that y|[0,b] is continuous and ℬ is a phase space.
1. Introduction
Fractional derivatives and integrals have been vastly used in different fields, facing a huge development especially during the last few decades (see, e.g., [1–9] and the references therein). The approaches based on fractional calculus establish models of engineering systems better than the ordinary derivatives approaches [1–6].
In particular, fractional differential equations as an important research branch of fractional calculus attracted much more attention (see, e.g., [10–20] and the references therein). Also varieties of schemes for numerical solutions of fractional differential equations are reported (see, e.g., [6, 21–23] and the references therein). We notice that some investigations have been done on the existence and uniqueness of solutions for fractional differential equations with delay (see, e.g., [24, 25] and the references therein).
Having all the aforementioned facts in mind, in this paper we study the existence and uniqueness of solutions for a class of delayed fractional differential equations, namely,
(1)ℒ(𝒟)y(t)=f(t,yt),t∈J=[0,b],y(t)=ϕ(t),t∈(-∞,0],
where ℒ(𝒟)=D0+α-tnD0+β,0<β<α<1,n is a positive integer, f:J×ℬ→ℝ is a given function satisfying some assumptions that will be specified later, ϕ∈ℬ with ϕ(0)=0, and ℬ is called a phase space that will be defined later. D0+α and D0+β are the standard Riemann-Liouville fractional derivatives. yt, which is an element ℬ, is defined as any function y on (-∞,b] as follows:
(2)yt(s)=y(t+s),s∈(-∞,0],t∈J.
Here yt(·) represents the preoperational state from time -∞ up to time t. We also consider the following nonlinear fractional differential equation:
(3)ℒ(𝒟){y(t)-g(t,yt)}=f(t,yt),t∈J,y(t)=ϕ(t),t∈(-∞,0],
where α, β, f, ϕ, and ℒ(𝒟) are as (1) and g:J×ℬ→ℝ is a given function which satisfies g(0,ϕ)=0.
The notion of the phase space ℬ plays an important role in the study of both qualitative and quantitative theories for functional differential equations. A common choice is a seminormed space satisfying suitable axioms, which was introduced by Hale and Kato [26].
Our approach is based on the Banach fixed-point theorem and on the nonlinear alternative of Leray-Schauder type [27, 28]. The organization of the paper is as follows.
In Section 2, we present some basic mathematical tools used in the paper. The main results are presented in Section 3. Section 4 is dedicated to our conclusions.
2. Preliminaries
In this section, we present some basic notations and properties which are used throughout this paper. First of all, we will explain the phase space ℬ introduced by Hale and Kato [26]. Let ℝ≤0=(-∞,0],ℝ≥0=[0,+∞),ℝ=(-∞,+∞), and let E be a Banach space with norm |·|E. Further, let ℬ be a linear space of functions mapping ℝ- into E with seminorm |·|ℬ having the following axioms,
If y:(-∞,σ+b)→E,b>0 is continuous on [σ,σ+b) and yσ∈ℬ, then yt∈ℬ and yt are continuous for any t∈[σ,σ+b).
There exist functions k(t)>0 and m(t)≥0 with the following properties. (i) k(t) is continuous for t∈ℝ≥0. (ii) m(t) is locally bounded for t∈ℝ≥0. (iii) For every function, y has the properties of (B1) and t∈[σ,σ+b), holds that |yt|ℬ≤k(t-σ)sup{|y(s)|E:σ≤s≤t}+m(t-σ)|yσ|ℬ.
There exists a positive constant L such that |ϕ(0)|E≤L|ϕ|ℬ for all ϕ∈ℬ.
The quotient space ℬ^:=ℬ/|·|ℬ is a Banach space.
We notice that in this paper, we select σ=0 and E=ℝ; thus (iii) can be converted to |yt|ℬ≤k(t)sup{|y(s)|E:0≤s≤t}+m(t)|y0|ℬ, for all t∈[0,b).
See [28] for examples of the phase space ℬ satisfying all axioms (B1)–(B4).
Let ℝ+=(0,+∞) and C0(ℝ+) be the space of all continuous real function on ℝ+. Consider also the space C0(ℝ)≥0 of all continuous real functions on ℝ≥0 which later identifies with the class of all f∈C0(ℝ+) such that limt→0+f(t)=f(0+)∈ℝ. By C(J,ℝ), we denote the Banach space of all continuous functions from J into ℝ with the norm ∥y∥∞:=sup{|y(t)|:t∈J}, where |·| is a suitable complete norm on ℝ.
The most common notation for αth order derivative of a real-valued function y(t), which is defined in an interval denoted by (a,b), is Daαy(t). Here, the negative value of α corresponds to the fractional integral.
Definition 1.
For a function y defined on an interval [a,b], the Riemann-Liouville fractional integral of y of order α>0 is defined by [1, 6]
(4)Ia+αy(t)=1Γ(α)∫at(t-s)α-1y(s)ds,t>a,
and the Riemann-Liouville fractional derivative of y(t) of order α>0 reads as
(5)Da+αy(t)=dndtn{Ia+n-αy(t)},n-1<α≤n.
Also, we denote Da+αy(t) as Daαy(t) and Ia+αy(t) as Iaαy(t). Further, D0+αy(t) and I0+αy(t) are referred to as Dαy(t) and Iαy(t), respectively. If the fractional derivative Daαy(t) is integrable, then we have [4, page 71]
(6)Iaα(Daβy(t))=Iaα-βy(t)-[Ia1-βy(t)]t=a(t-a)α-1Γ(α),0<β≤α<1.
If y is continuous on [a,b], then Daαy(t) is integrable, I1-βy(t)|t=a=0, and
(7)Iaα(Daβy(t))=Iaα-βy(t),0<β≤α<1.
Proposition 2.
Let y be continuous on [0,b] and n a nonnegative integer, then
(8)(i)Iα(tny(t))=∑k=0n(-αk)[Dktn][Iα+ky(t)]=∑k=0n(-αk)n!tn-k(n-k)!Iα+ky(t),(9)(ii)Iα(tnDβy(t))=∑k=0n(-αk)n!tn-k(n-k)!Iα-β+ky(t),
where
(10)(-αk)=(-1)kΓ(α+1)k!Γ(α)=(-1)k(α+k-1k)=Γ(1-α)Γ(k+1)Γ(1-α-k).
Proof.
(i) can be found in [6, page 53], and (ii) is an immediate consequence of (7), and (i).
Lemma 3 (see [29]).
Let v:[0,b]→[0,∞) be a real function and w(·) a nonnegative, locally integrable function on [0,b]. If there exist positive constants a and α∈(0,1) such that v(t)≤w(t)+a∫0t(t-s)-αv(s)ds, then there exists a constant K=K(α) such that v(t)≤w(t)+Ka∫0tw(s)(t-s)-αds, for all t∈[0,b].
In this paper we use the alternative Leray-Schauder’s theorem and Banach’s contraction principle for getting the main results. These theorems can be found in [27, 28].
3. Existence and Uniqueness
In this section, we prove the existence results for (1) and (3) by using the alternative of Leray-Schauder’s theorem. Further, our results for the unique solution is based on the Banach contraction principle. Let us start by defining what we mean by a solution of (1). Let the space
(11)Ω={y:(-∞,b]→ℝ:y∣(-∞,0]∈ℬandy|[0,b]iscontinuous}.
A function y∈Ω is said to be a solution of (1) if y satisfies (1).
For the existence results on (1), we need the following lemma.
Lemma 4.
Equation (1) is equivalent to the Volterra integral equation
(12)y(t)=∑k=0n(-αk)n!tn-k(n-k)!Iα-β+ky(t)+Iαf(t,yt),t∈J.
Proof.
The proof is an immediate consequence of Proposition 2.
To study the existence and uniqueness of solutions for (1), we transform (1) into a fixed-point problem. Consider the operator P:Ω→Ω defined by
(13)Py(t)={ℒ(I)y(t)+Iαf(t,yt),t∈[0,b],ϕ(t),t∈(-∞,0],
where,
(14)ℒ(I)=∑k=0n(-αk)n!tn-k(n-k)!Iα-β+k.
Let x(·):(-∞,b]→ℝ be the function defined as
(15)x(t)={0,ift∈[0,b],ϕ(t),ift∈(-∞,0].
Then, we get x0=ϕ. For each z∈C([0,b],ℝ) with z(0)=0, we denote by z- the function defined as follows:
(16)z-(t)={z(t),ift∈[0,b],0,ift∈(-∞,0].
If y(·) satisfies the integral equation y(t)=ℒ(I)y(t)+Iαf(t,yt), then we can decompose y(·) as y(t)=z-(t)+x(t), -∞<t≤b, which implies yt=z-t+xt for every 0≤t≤b, and the function z(·) satisfies
(17)z(t)=ℒ(I)z(t)+Iαf(t,z-t+xt),
set C0={z∈C([0,b],ℝ):z(0)=0}, and let ∥·∥b be the seminorm in C0 defined by ∥z∥b=∥z0∥ℬ+sup{|z(t)|:0≤t≤b}=sup{|z(t)|:0≤t≤b},z∈C0. C0 is a Banach space with norm ∥·∥b. Let the operator F:C0→C0 be defined by
(18)Fz(t)=ℒ(I)z(t)+Iαf(t,z-t+xt),
where t∈[0,b]. The operator P has a fixed point equivalent to F that has a fixed point too.
Theorem 5.
Assume that f is a continuous function, and there exist p,q∈C(J,ℝ+) such that |f(t,u)|≤p(t)+q(t)∥u∥ℬ,t∈J,u∈ℬ. Then, (1) has at least one solution on (-∞,b].
Proof.
It is enough to show that the operator F:C0→C0 defined as (18) satisfies the following: (i) F is continuous, (ii) F maps bounded sets into bounded sets in C0, (iii) F maps bounded sets into equicontinuous sets of C0, and (iv) F is completely continuous.
Let {zn} converges to z in C0, then
(19)∥Fzn(t)-Fz(t)∥≤∑k=0n|(-αk)|n!tn-k(n-k)!Iα-β+k|zn(t)-z(t)|+Iα|f(t,(z-n)t+xt)-f(t,z-t+xt)|≤∑k=0n|(-αk)|n!bn-k∥zn-z∥(n-k)!Γ(α-β+k+1)+bα∥f(t,(z-n)t+xt)-f(t,z-t+xt)∥Γ(α+1).
Hence, ∥Fzn(t)-Fz(t)∥→0 as zn→z, and thus f is continuous.
For any λ>0, let ℬλ={z∈C0:∥z∥b≤λ} be a bounded set. We show that there exists a positive constant μ such that ∥Fz∥∞≤μ. Let z∈ℬλ, since f is a continuous function, we have for each t∈[0,b],
(20)|Fz(t)|≤∑k=0n|(-αk)|n!tn-k(n-k)!Γ(α-β+k)×∫0b(t-s)α-β+k-1z(s)+1Γ(α)∫0t(t-s)α-1f(s,z-s+xs)ds≤∑k=0n|(-αk)|n!bn+α-β(n-k)!Γ(α-β+k+1)∥z∥b+1Γ(α)×∫0t(t-s)α-1[p(s)+q(s)∥z-s+xs∥ℬ]ds≤∑k=0n|(-αk)|n!bn+α-β(n-k)!Γ(α-β+k+1)∥z∥b+bα∥p∥∞Γ(α+1)+bα∥q∥∞Γ(α+1){∥z-s∥ℬ+∥xs∥ℬ}≤∑k=0n|(-αk)|n!bn+α-β(n-k)!Γ(α-β+k+1)∥z∥b+bα∥p∥∞Γ(α+1)+kbλ+mb∥ϕ∥ℬ:=μ,
where mb=sup{|m(t)|:t∈[0,b]}, and kb=sup{|k(t)|:t∈[0,b]}. Hence, we obtain ∥Fz∥∞≤μ.
Let t1,t2∈[0,b] and t1<t2. Let ℬλ be a bounded set of C0 as in (ii) and z∈ℬλ, then given ϵ>0 choose
(21)δ=min{12Λ1ϵ1/α,12(n+1)Λ2ϵ1/(α-β+k):k=0,1,…,n12(n+1)Λ2},
where
(22)Λ2=Λ1=2∥p∥∞+Λ∥q∥∞Γ(α+1),Λ2=∑k=0n2|(-αnk)|k!bn-k∥z∥b(n-k)!Γ(α-β+k+1),
and Λ=kbλ+mb∥ϕ∥ℬ. If |t2-t1|<δ, then
(23)|Fz(t2)-Fz(t1)|≤∑k=0n|(-αnk)|k!bn-k(n-k)!Γ(α-β+k)∥z∥b×|∫0t1{(t2-s)α-β+k-1-(t1-s)α-β+k-1}ds+∫t1t2(t2-s)α-β+k-1ds|+1Γ(α)|∫0t1{(t2-s)α-1-(t1-s)α-1}f(s,z-s+xx)ds+∫t1t2(t2-s)α-1f(s,z-s+xx)ds|≤∑k=0n2|(-αnk)|k!bn-k(n-k)!Γ(α-β+k+1)∥z∥b(t2-t1)α-β+k+∥p∥∞+Λ∥q∥∞Γ(α+1){∫0t1{(t2-s)α-1-(t1-s)α-1}ds+∥p∥∞+Λ∥q∥∞Γ(α+1)+∫t1t2(t2-s)α-1ds}≤∑k=0n2|(-αnk)|k!bn-k(n-k)!Γ(α-β+k+1)∥z∥b(t2-t1)α-β+k+2∥p∥∞+Λ∥q∥∞Γ(α+1)(t2-t1)α=Λ2δα-β+k+Λ1δα<ϵ2+ϵ2=ϵ,
where ∥z-s+xs∥ℬ≤∥z-s∥ℬ+∥xs∥ℬ≤kbλ+mb∥ϕ∥ℬ:=Λ. Hence, F(ℬλ) is equicontinuous.
It is an immediate consequence from (i)–(iii), together with the Arzela-Ascoli theorem.
We show in the following that there exists an open set U⊆C0 with z≠γF(z) for γ∈(0,1) and z∈∂U. Let z∈C0 and z=γF(z) for some 0<γ<1. Then, for each t∈[0,b], we have z(t)=λ{ℒ(I)z(t)+Iαf(t,z-t+xt)}. It follows by assumption of the theorem
(24)|z(t)|≤∑k=0n|(-αnk)|k!bn-k(n-k)!Γ(α-β+k)∫0t(t-s)α-β+k-1|z(s)|ds+1Γ(α)∫0t(t-s)α-1|f(s,z-s+xx)|ds≤∑k=0n|(-αnk)|k!bn-k∥z∥b(n-k)!Γ(α-β+k+1)+1Γ(α)∫0t(t-s)α-1q(s)∥z-s+xs∥ℬds+bα∥p∥∞Γ(α+1).
On other hand, we have
(25)∥z-s+xs∥B≤∥z-s∥ℬ+∥xs∥ℬ≤k(t)sup{|z(s)|:0≤s≤t}+m(t)∥z0∥ℬ+k(t)sup{|x(s)|:0≤s≤t}+m(t)∥x0∥ℬ≤kbsup{|z(s)|:0≤t≤t}+mb∥ϕ∥ℬ.
If we let δ(t) the right-hand side of (25), then ∥z-s+xs∥ℬ≤δ(t) and, therefore,
(26)|z(t)|≤∑k=0n|(-αnk)|k!bn-k∥z∥b(n-k)!Γ(α-β+k+1)+1Γ(α)∫0t(t-s)α-1q(s)δ(s)ds+bα∥p∥∞Γ(α+1).
Using the aforementioned inequality and the definition of δ, we get
(27)δ(t)≤∑k=0n|(-αnk)|k!bn-k∥z∥bkb(n-k)!Γ(α-β+k+1)+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1)+kb∥q∥∞Γ(α)×∫0t(t-s)α-1δ(s)ds.
Then, using Lemma 3, there exists a constant Δ such that
(28)|δ(t)|≤12kbΛ2+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1)+Δkb∥q∥∞Γ(α)∫0t(t-s)α-1Rds,
where Λ2 is mentioned in (22), and
(29)R=12kbΛ2+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1).
Hence,
(30)∥δ∥∞≤R+RΔbαkb∥q∥∞Γ(α+1):=M~,
and then ∥z∥∞≤Λ2+M~∥Iαq∥∞+bα∥p∥∞/Γ(α+1). Therefore,
(31)∥z∥∞≤M~∥Iαq∥∞+bα∥p∥∞/Γ(α+1)1-Λ2:=Δ*.
Set U={z∈C0:∥z∥b<Δ*+1}. Then, F:U¯→C0 is continuous and completely continuous. From the choice of U, there is no z∈∂U such that z=γF(z), for γ∈(0,1); therefore, by the nonlinear alternative of the Leray-Schauder theorem, the proof is complete.
Theorem 6.
Let f:J×B→ℝ be a continuous function. If there exists a positive constant l such that |f(t,u)-f(t,v)|≤l∥u-v∥ℬ,t∈J,u,v∈ℬ, and 0<T+lkbbα/Γ(α+1):=L<1 then (1) has a unique solution in the interval (-∞,b], where,
(32)T=∑k=0n|(-αnk)|k!bn-k(n-k)!Γ(α-β+k+1).
Proof.
The solution of (1) is equivalent to the solution of the integral equation (17). Hence, it is enough to show that the operator F:C0→C0, satisfies the Banach fixed-point theorem. Consider u,v∈C0 and for each t∈[0,b], we have
(33)|F(z)(t)-F(u)(t)|≤T∥u-v∥b+1Γ(α)∫0t(t-s)α-1l∥u-s-v-s∥ℬds≤T∥u-v∥b+lΓ(α)∫0t(t-s)α-1∥u-s-v-s∥ℬds≤T∥u-v∥b+lΓ(α)∫0t(t-s)α-1×kbsup∥u(s)-v(s)∥ds≤{T+lkbΓ(α)∫0t(t-s)α-1lds}|u-v|b≤{T+lkbbαΓ(α+1)}∥u-v∥b=L∥u-v∥b.
Hence, ∥F(z)-F(v)∥b≤L∥z-z*∥b, and then F is a contraction. Therefore, F has a unique fixed point by Banach’s contraction principle.
Theorem 7.
Let f:J×ℬ→ℝ be a continuous function, and let the following assumptions hold.
There exist p,q∈C(J,ℛ≥0) such that |f(t,u)|≤p(t)+q(t)∥u∥ℬ for each t∈J,u∈ℬ and and ∥Iαp∥<+∞.
The function g is continuous and completely continuous. For any bounded set 𝒟 in Ω, the set {t→g(t,yt):y∈𝒟} is equicontinuous in C([0,b],ℝ). There exist positive constants d1 and d2 such that |g(t,u)|≤d1∥u∥ℬ+d2 for each t∈[0,b] and u∈ℬ.
If kbd1∈(0,1), then (3) has at least one solution on (-∞,b], where kb=sup{|k(t)|:t∈[0,b]}.
Proof.
Consider the operator P*:Ω→Ω defined by
(34)P*(y)(t)={ℒ(I)y(t)+Iαf(t,yt)+g(t,yt),t∈[0,b],ϕ(t),t∈(-∞,0],
where
(35)ℒ(I)=∑k=0n(-αk)n!tn-k(n-k)!Iα-β+k.
In analog to Theorem 5, we consider the operator F*:C0→C0 defined by
(36)F*z(t)=ℒ(I)z(t)+Iαf(t,z-t+xt)+g(t,z-t+xt).
By using (H2) and Theorem 5, the operator F* is continuous and completely continuous. Now, it is sufficient to show that there exists an open set U*⊆C0 with z≠λF*(z) for γ∈(0,1) and z∈∂U*.
Let z∈C0 and z=γF*(z) for some γ∈(0,1). Then, for each t∈[0,b],z(t)=γ[g(t,z-t+xt)+ℒ(I)z(t)+Iαf(t,z-t+xt)]. Hence,
(37)|z(t)|≤d1∥z-t+xt∥ℬ+d2+∑k=0n|(-αk)|k!bn-k(n-k)!Γ(α-β+k+1)∥z∥b+bαn∥p∥∞Γ(α+1)+1Γ(α)∫0t(t-s)αn-1q(s)∥z-s+xs∥ℬds,≤d1δ(t)+d2+bα∥p∥∞Γ(α+1)+1Γ(α)∫0t(t-s)α-1q(s)δ(s)ds+∑k=0n|(-αk)|k!bn-k(n-k)!Γ(α-β+k+1)∥z∥b,
where δ(t) is named the in right-hand side of (25) such that ∥z-s-xs∥≤δ(t). Since 0<kbd1<1, if we let T*=∑k=0n(|(-αk)|k!bn-k∥z∥bkb/(n-k)!Γ(α-β+k+1)), then
(38)δ(t)≤kbd1δ(t)+kbd2+mb∥ϕ∥ℬ+T*+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1)+kb∥q∥∞Γ(α)∫0t(t-s)α-1δ(s)dsδ(t)≤11-kbd1{∫0tkbd2+mb∥ϕ∥ℬ+T*+mb∥ϕ∥ℬ2+kbbα∥p∥∞Γ(α+1)+kb∥q∥∞Γ(α)2×∫0t(t-s)α-1δ(s)ds}.
Then, using Lemma 3, there exists a constant Δ* such that
(39)δ(t)≤kbd1δ(t)+kbd2+mb∥ϕ∥ℬδ(t)+T*+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1)δ(t)+kb∥q∥∞Γ(α)∫0t(t-s)α-1δ(s)dsδ(t)≤11-kbd1δ(t)×{kbbα∥p∥∞Γ(α+1)kbd2+mb∥ϕ∥ℬ+T*+mb∥ϕ∥ℬ+kbbα∥p∥∞Γ(α+1)δ(t)+Δ*kb∥q∥∞Γ(α)∫0t(t-s)α-1δ(s)dskbbα∥p∥∞Γ(α+1)},
and, therefore, ∥w∥∞≤R+RΔ*kb∥q*∥∞/Γ(α+1):=L*, where ∥q*∥∞=∥q∥∞/(1-kbd1) and R=1/(1-kbd1)[kbd2+mb∥ϕ∥ℬ+(kbbα∥p∥∞)/Γ(α+1)+T*]. Then,
(40)∥z∥∞≤d1L*+d2+bα∥p∥∞Γ(α+1)+L∥Iαq∥∞+T*,
and, hence,
(41)∥z∥∞≤d1L*+d2+bα∥p∥∞/Γ(α+1)+L*∥Iαq∥∞1-∥z∥∞T*:=M*.
Set U*={z∈C0:∥z∥b<M*+1}. From the choice of U*, there is no z∈∂U* such that z=γF*(z) for γ∈(0,1). As a consequence of the nonlinear alternative of the Leray-Schauder theorem, we deduce that F* has a fixed-point z* in U*, which is a solution of (3).
The unique solution of (3), under some conditions, is studied in the following theorem which is the result of the Banach contraction mapping.
Theorem 8.
Let f:J×ℬ→ℝ be a continuous function, and there exist positive constants l,μ, such that
(42)|f(t,u)-f(t,v)|≤l∥u-v∥ℬ,|g(t,u)-g(t,v)|≤μ∥u-v∥ℬ,
where t∈J and u,v∈ℬ. Then, (3) with the following conditions has a unique solution in the interval (-∞,b](43)0<T+lkbbαΓ(α+1)<1,0<kbμ+T+kblbαΓ(α+1)<1,
such that T is defined in Theorem 6.
Proof.
The proof is a similar process Theorem 6.
4. Conclusions
In this paper, the existence and the uniqueness of solutions for the nonlinear fractional differential equations with infinite delay comprising standard Riemann-Liouville derivatives have been discussed in the phase space. Leray-Schauder’s alternative theorem and the Banach contraction principle were used to prove the obtained results. Further generalizations can be developed to some other class of fractional differential equations such as ℒ(D)y(t)=f(t,yt), where ℒ(D)=Dαn-∑j=1n-1pj(t)Dαn-j,0<α1<⋯<αn<1,pj(t)=∑k=0Njajktk, and Nj is nonnegative integer.
Acknowledgments
The authors would like to thank the referee for helpful comments and suggestions. This paper was funded by King Abdulaziz University, under Grant no. (130-1-1433/HiCi). The authors, therefore, acknowledge technical and financial support of KAU.
KilbasA. A.SrivastavaH. M.TrujilloJ. J.2006204Amsterdam, The NetherlandsElsevier Science B.V.xvi+523North-Holland Mathematics StudiesMR221807310.1016/S0304-0208(06)80001-0ZBL1206.26007MillerK. S.RossB.1993New York, NY, SUAJohn Wiley & Sonsxvi+366A Wiley-Interscience PublicationMR1219954ZBL0943.82582OldhamK. B.SpanierJ.1974New York, NY, USAAcademic Pressxiii+234MR0361633ZBL0292.26011PodlubnyI.1999198San Diego, Calif, USAAcademic Pressxxiv+340Mathematics in Science and EngineeringMR1658022ZBL1056.93542SamkoS. G.KilbasA. A.MarichevO. I.1993Yverdon, SwitzerlandGordon and Breach Sciencexxxvi+976MR1347689ZBL0924.44003BaleanuD.DiethelmK.ScalasE.TrujilloJ. J.20123Hackensack, NJ, USAWorld Scientificxxiv+400Series on Complexity, Nonlinearity and Chaos10.1142/9789814355216MR2894576ZBL1251.93029GlockleW. G.NonnenmacherT. F.A fractional calculus approach of self-similar protein dynamics1995684653BaleanuD.GünvencZ. B. G.MachadoJ. A. T.2010New York, NY, USASpringerxii+53110.1007/978-90-481-3293-5MR2605606ZBL1237.82041DiethelmK.FreedA. D.KeilF.MackensW.VossH.WertherJ.On the solution of fractional order differential equations used in the modeling of viscoplasticity1999Heidelberg, GermanySpringer217224BabakhaniA.Daftardar-GejjiV.Existence of positive solutions of nonlinear fractional differential equations2003278243444210.1016/S0022-247X(02)00716-3MR1974017ZBL1027.34003YuC.GaoG.Existence of fractional differential equations20053101262910.1016/j.jmaa.2004.12.015MR2160670ZBL1088.34501BabakhaniA.Daftardar-GejjiV.Existence of positive solutions for N-term non-autonomous fractional differential equations20059219320610.1007/s11117-005-2715-xMR2189743BăleanuD.MustafaO. G.AgarwalR. P.Asymptotically linear solutions for some linear fractional differential equations20102010886513910.1155/2010/865139MR2746005ZBL1210.34005CabadaA.WangG.Positive solutions of nonlinear fractional differential equations with integral boundary value conditions2012389140341110.1016/j.jmaa.2011.11.065MR2876506ZBL1232.34010AhmadB.NietoJ. J.AlsaediA.El-ShahedM.A study of nonlinear Langevin equation involving two fractional orders in different intervals201213259960610.1016/j.nonrwa.2011.07.052MR2846866ZBL1238.34008BhalekarS.Daftardar-GejjiV.BaleanuD.MaginR.Fractional Bloch equation with delay20116151355136510.1016/j.camwa.2010.12.079MR2773409ZBL1217.34123AgarwalR. P.AhmadB.Existence theory for anti-periodic boundary value problems of fractional differential equations and inclusions20116231200121410.1016/j.camwa.2011.03.001MR2824708ZBL1228.34009AhmadB.NietoJ. J.AlsaediA.A nonlocal three-point inclusion problem of Langevin equation with two different fractional orders20122012, article 54BaleanuD.TrujilloJ. J.On exact solutions of a class of fractional Euler-Lagrange equations200852433133510.1007/s11071-007-9281-7MR2401415ZBL1170.70328DiethelmK.FordN. J.Analysis of fractional differential equations2002265222924810.1006/jmaa.2000.7194MR1876137ZBL1014.34003DiethelmK.An algorithm for the numerical solution of differential equations of fractional order1997516MR1447831ZBL0890.65071OdibatZ.MomaniS.Numerical methods for nonlinear partial differential equations of fractional order2008322839DiethelmK.WalzG.Numerical solution of fractional order differential equations by extrapolation1997163-423125310.1023/A:1019147432240MR1617164ZBL0926.65070MaraabaT.BaleanuD.JaradF.Existence and uniqueness theorem for a class of delay differential equations with left and right Caputo fractional derivatives20084981108350710.1063/1.2970709MR2440705ZBL1152.81550YeH.DingY.GaoJ.The existence of a positive solution of a fractional differential equation with delay200711341350HaleJ. K.KatoJ.Phase space for retarded equations with infinite delay19782111141MR0492721ZBL0383.34055GranasA.DugundjiJ.2003New York, NY, USASpringerxvi+690Springer Monographs in MathematicsMR1987179GolbergR. R.1970New Delhi, IndiaOxford and IBH Publishing CompanyHenryD.1989Berlin, GermanySpringer