Agarwal et al. Advances in Difference Equations (2015) 2015:112 DOI 10.1186/s13662-015-0451-5

RESEARCH

Open Access

Nonlinear fractional differential equations in nonreflexive Banach spaces and fractional calculus Ravi P Agarwal1,2* , Vasile Lupulescu3 , Donal O’Regan2,4 and Ghaus ur Rahman5 *

Correspondence: [email protected] 1 Department of Mathematics, Texas A&M University-Kingsville, Kingsville, TX 78363, USA 2 Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah, 21589, Saudi Arabia Full list of author information is available at the end of the article

Abstract The aim of this paper is to correct some ambiguities and inaccuracies in Agarwal et al. (Commun. Nonlinear Sci. Numer. Simul. 20(1):59-73, 2015; Adv. Differ. Equ. 2013:302, 2013, doi:10.1186/1687-1847-2013-302) and to present new ideas and approaches for fractional calculus and fractional differential equations in nonreflexive Banach spaces.

1 Introduction One of the sections, Section , of our paper [] contains a number of ambiguities (and inaccuracies) which we correct here. The notion of pseudo-solution in [] is not adequately defined and assumption (h) in Theorem . is strong. Also, there is some ambiguity regarding the use of the space C(T, E) of all continuous functions from T into E with its weak topology σ (C(T, E), C(T, E)∗ ) and the space Cw (T, E) of all weakly continuous functions from T into Ew endowed with the topology of weak uniform convergence. Parts of Corollaries .-. are no longer valid in their current form. Similar comments also apply to []. In [] the authors developed fractional calculus for vector-valued functions using the weak Riemann integral and they established the existence of weak solutions for a class of fractional differential equations with fractional weak derivatives. In this paper we present new ideas in fractional calculus and we present a new approach to establishing existence to some fractional differential equations in nonreflexive Banach spaces. References [–], and [] were helpful in presenting these new ideas. 2 Preliminaries In the following we outline some aspects of fractional calculus in a nonreflexive Banach space. This subject has been treated extensively in [, ]. Let E be a Banach space with norm  ·  and let E∗ be the topological dual of E. If x∗ ∈ E∗ , then its value on an element x ∈ E will be denoted by x∗ , x. The space E endowed with the weak topology σ (E, E∗ ) will be denoted by Ew . Consider an interval T = [, b] of R, the set of real numbers, endowed with the Lebesgue σ -algebra L(T) and the Lebesgue measure λ. A function x(·) : T → E is said to be strongly measurable on T if there exists a sequence of simple functions xn (·) : T → E such that limn→∞ xn (t) = x(t) for a.e. t ∈ T. Also, a function x(·) : T → E is said to be weakly measurable (or scalarly measurable) on T if, for every x∗ ∈ E∗ , the real valued function t → x∗ , x(t) is Lebesgue measurable on T. © 2015 Agarwal et al.; licensee Springer. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 2 of 18

We denote by Lp (T) the space of all real measurable functions f : T → R, whose absolute value raised to the pth power has finite integral, or equivalently, that 

 p f (t) dt

f p :=

 p

< ∞,

T

where  ≤ p < ∞. Moreover, by L∞ (T) we denote the space of all measurable and essential bounded real functions defined on T. Let C(T, E) denote the space of all strong continuous functions y(·) : T → E, endowed with the supremum norm y(·)c = supt∈T y(t). Also, we consider the space C(T, E) with its weak topology σ (C(T, E), C(T, E)∗ ). It is well known that (see [, ])   C(T, E)∗ = M T, E∗ , where M(T, E∗ ) is the space of all bounded regular vector measures from B (T) into E∗ which are of bounded variation. Here, B (T) denotes the σ -algebra of Borel measurable subsets of T. Therefore, a sequence {yn (·)}n≥ converges weakly to y(·) in C(T, E) if and only if  m(·), yn (·) – y(·) → 

as n → ∞,

()

for all m(·) ∈ M(T, E∗ ). In [], Lemma , it is shown that a sequence {yn (·)}n≥ converges weakly to y(·) in C(T , E) if and only if yn (t) tends weakly to y(t) for each t ∈ T. Let Cw (T, E) denote the space of all weakly continuous functions from T into Ew endowed with the topology of weak uniform convergence. A set N ∈ L(T) is called a null set if λ(N) = . A function x(·) : T → E is said to be pseudo-differentiable on T to a function y(·) : T → E if, for every x∗ ∈ E∗ , there exists a null set N(x∗ ) ∈ L(T) such that the real function t → x∗ , x(t) is differentiable on T  N(x∗ ) and  d ∗ x , x(t) = x∗ , y(t) , dt

  t ∈ T  N x∗ .

()

The function y(·) is called a pseudo-derivative of x(·) and it will be denoted by x p (·) or by dp x(·). A pseudo-derivative dt

x p (·) of a pseudo-differentiable function x(·) : T → E is weakly

measurable on T (see []). We recall that a function x(·) : T → E is said to be weakly differentiable on T if there exists a function x w (·) : T → E such that

  x(t + h) – x(t ) = x∗ , x w (t ) , lim x∗ , h→ h for every x∗ ∈ E∗ . If it exists, x w (·) is uniquely determined and it is called the weak derivative of x(·) on T. Obviously, if x(·) : T → E is a weakly differentiable function on T, then the real function t → x∗ , x(t) is differentiable on T. Moreover, in this case we have  d ∗ x , x(t) = x∗ , x w (t) , dt

t ∈ T,

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 3 of 18

for every x∗ ∈ E∗ . It is easy to see that, if x(·) : T → E is a function a.e. weakly differentiable on T, then x(·) is pseudo-differentiable on T and x p (·) = x w (·) a.e. on T. The concept of a Bochner integral and a Pettis integral are well known [–]. We recall that a weakly measurable function x(·) : T → E is said to be Pettis integrable on T if (a) x(·) is scalarly integrable; that is, for every x∗ ∈ E∗ , the real function t → x∗ , x(t) is Lebesgue integrable on T; (b) for every set A ∈ L(T), there exists an element xA ∈ E such that 

x∗ , xA =



 ∗ x , x(s) ds,

()

A

for every x∗ ∈ E∗ . The element xA ∈ E is called the Pettis integral on A and it will be denoted by A x(s) ds. It is easy to show that a Bochner integrable function x(·) : T → E is Pettis integrable and both integrals of x(·) are equal on each Lebesgue measurable subset A of T ([], Proposition ..). The best result for a descriptive definition of the Pettis integral is that given by Pettis in []. Proposition  Let x(·) : T → E be a weakly measurable function. (a) If x(·) is Pettis integrable on T, then the indefinite Pettis integral 

t

y(t) :=

x(s) ds,

t∈T



is AC on T and x(·) is a pseudo-derivative of y(·). (b) If y(·) : T → E is an AC function on T and it has a pseudo-derivative x(·) on T, then x(·) is Pettis integrable on T and 

t

y(t) = y() +

x(s) ds,

t ∈ T.



It is well known that the Pettis integrals of two strongly measurable functions x(·) : T → E and y(·) : T → E coincide over every Lebesgue measurable set in T if and only if x(·) = y(·) a.e. on T ([], Theorem .). Since a pseudo-derivative of a pseudo-differentiable function x(·) : T → E is not unique (see []) and two pseudo-derivatives of x(·) need not be a.e. equal, the concept of weakly equivalence plays an important role in the following. Two weak measurable functions x(·) : T → E and y(·) : T → E are said to be weakly equivalent on T if, for every x∗ ∈ E∗ , we have x∗ , x(t) = x∗ , y(t) for a.e. t ∈ T. In the following, if two weak measurable functions x(·) : T → E and y(·) : T → E are weakly equivalent on T, then we will write x(·)  y(·) or x(t)  y(t), t ∈ T. Proposition  ([]) A weakly measurable function x(·) : T → E is Pettis integrable on T and x∗ , x(·) ∈ L∞ (T), for every x∗ ∈ E∗ , if and only if the function t → ϕ(t)x(t) is Pettis integrable on T, for every ϕ(·) ∈ L (T). Let us denote by P∞ (T, E) the space of all weakly measurable and Pettis integrable functions x(·) : T → E with the property that x∗ , x(·) ∈ L∞ (T), for every x∗ ∈ E∗ . Since for each

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 4 of 18

t ∈ T the real valued function s → (t – s)α– is Lebesgue integrable on [, t], the fractional Pettis integral 

t

I α x(t) := 

(t – s)α– x(s) ds, (α)

t ∈ T,

exists, for every function x(·) ∈ P∞ (T, E), as a function from T into E (see []). Moreover, we have  ∗ α x , I x(t) =



t 

(t – s)α–  ∗ x , x(s) ds, (α)

t ∈ T,

for every x∗ ∈ E∗ , and the real function t → x∗ , I α x(t) is continuous (in fact, bounded and uniformly continuous on T if T = R) on T, for every x∗ ∈ E∗ ([], Proposition ..). In the following, consider α ∈ (, ) and for a given function x(·) ∈ P∞ (T, E) we also denote by x–α (t) the fractional Pettis integral  I

–α

x(t) = 

t

(t – s)–α x(s) ds, ( – α)

t ∈ T.

Lemma  ([], Lemma .) If x(·), y(·) ∈ P∞ (T, E) are weakly equivalent on T, then I α x(t) = I α y(t) on T. Lemma  ([, ]) The fractional Pettis integral is a linear operator from P∞ (T, E) into P∞ (T, E). Moreover, if x(·) ∈ P∞ (T, E), then for α, β >  we have (a) I α I β x(t) = I α+β x(t), t ∈ T; (b) limα→ I α x(t) = I  x(t) = x(t) – x() weakly uniformly on T; (c) limα→ I α x(t) = x(t) weakly on T. If y(·) : T → E is a pseudo-differentiable function on T with a pseudo-derivative x(·) ∈ P (T, E), then the fractional Pettis integral I –α x(t) exists on T. The fractional Pettis integral I –α x(·) is called a fractional pseudo-derivative of y(·) on T and it will be denoted by Dαp y(·); that is, ∞

Dαp y(t) = I –α x(t),

t ∈ T.

()

Remark  If x(·), x(·) ∈ P∞ (T, E) are two pseudo-derivatives of y(·) : T → E, then x(·) 

x(t) on T, and so Dαp y(·) does not x(·) on T. Thus, Lemma  implies that I –α x(t) = I –α depend on the choice of a pseudo-derivatives of the function y(·). Therefore, we can write () as Dαp y(t) = I –α y p (t),

t ∈ T,

()

where y p (·) is a given pseudo-derivatives of y(·). We recall that a function x(·) : T → E is said to be weakly absolutely continuous (wAC, for short) on T if, for every x∗ ∈ E∗ , the real valued function t → x∗ , x(t) is absolutely continuous on T.

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 5 of 18

Lemma  ([]) If y(·) ∈ P∞ (T, E) is a pseudo-differentiable function on T with a pseudoderivative x(·) ∈ P∞ (T, E), then the function 

t

y–α (t) := 

(t – s)–α y(s) ds, ( – α)

t ∈ T,

is wAC and it has a pseudo-derivative

dp y (·) ∈ P∞ (T, E) dt –α

dp t –α y–α (t)  y() + I –α x(t) on T. dt ( – α)

such that

()

Remark  Relation () can be written as dp t –α y–α (t) – y() dt ( – α)

Dαp y(t) 

on T.

()

Note () suggests us that we can extend the definition of the fractional pseudo-derivative for functions y(·) ∈ P∞ (T, E) for which the function t → y–α (t) is pseudo-differentiable d d on T. If dtp y–α (t) exists on T, then dtp y–α (t) will be called the Riemann-Liouville fracd tional pseudo-derivative of y(·) and it will be denoted by Dpα y(·); that is, Dpα y(·) = dtp y–α (·). Usually, Dαp y(·) is called the Caputo fractional pseudo-derivative of y(·). Relation () can be written as Dαp y(t)  Dpα y(t) –

t –α y() on T. ( – α)

()

Therefore, the Caputo fractional pseudo-derivative Dαp y(·) exists together with the Riemann-Liouville fractional pseudo-derivative Dpα y(·) and they satisfy (). It is easy to see that if y() = , then Dαp y(t)  Dpα y(t) on T.

()

Remark  Let y(·) : T → E be a pseudo-differentiable function with a pseudo-derivative y p (·) ∈ P∞ (T, E). Then from Lemma  we find that the function 

t

yα (t) := 

(t – s)α– y(s) ds, (α)

t ∈ T,

is wAC and has a pseudo-derivative

dp y (t) dt α

dp t α– yα (t) – y() dt (α)

on T.

D–α p y(t) 

Lemma  Let α, β ∈ (, ). (a) If y(·) ∈ P∞ (T, E), then Dαp I α y(t) = y(t),

t ∈ T.

such that

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 6 of 18

(b) If y(·) ∈ P∞ (T, E) and y–α (·) is pseudo-differentiable with a pseudo-derivative dp y (·) ∈ P∞ (T, E), then dt –α I α Dαp y(t) = y(t) – y(),

t ∈ T.

Proof (a) Indeed, since y(·) ∈ P∞ (T, E), then t → x∗ , y(t) is essentially bounded on T, for every x∗ ∈ E∗ . Hence we have  ∗ α        x , I y(t)  = I α x∗ , y(t)  ≤ M x∗

tα , ( + α)

t ∈ T,

where M(x∗ ) = ess supt∈T |x∗ , y(t)| < ∞, x∗ ∈ E∗ . Since the real function t → x∗ , I α y(t) is continuous on T, it follows that x∗ , I α y() = , for every x∗ ∈ E∗ , and thus I α y() = . β β Then by Remark  we have Dp I α y(t) = Dp I α y(t), and so by Lemma  and Proposition  we have Dαp I α y(t) = Dpα I α y(t) =

dp –α α dp dp I I y(t) = I  y(t) = dt dt dt



t

y(s) ds = y(t),

t ∈ T.



(b) By Lemma  and Proposition  we have I α Dαp y(t) = I α I –α y p (t) = I  y p (t) =



t 

y p (s) ds = y(t) – y(),

t ∈ T.



Lemma  Let y(·) : T → E be a pseudo-differentiable function on T with y p (·) ∈ P∞ (T, E) and  < α ≤ β < . Then we have (a) I α Dβp y(t) = Dβ–α () p y(t) on T. (b) If y() = , then Dβp I α y(t) = Dβ–α p y(t) on T

()

I β Dαp y(t) = I β–α y(t)

()

and on T.

Proof If y(·) : T → E is a pseudo-differentiable function on T, then by Lemma  we have I α Dβp y(t) = I α I –β y p (t) = I –(β–α) y p (t) = Dβ–α p y(t),

t ∈ T.

If y() = , then by Remark  and () we have Dβp I α y(t) = I –β

dp β–α yα (t) = I –β D–α p y(t) = Dp y(t), dt

t ∈ T.

Also, since y() = , then by Lemma  and Proposition  we have I β Dαp y(t) = I β I –α y p (t) = I β–α+ y p (t) = I β–α I  y p (t) = I β–α y(t),

t ∈ T.



Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 7 of 18

3 Differential equations with fractional pseudo-derivatives The existence of weak solutions or pseudo-solutions for ordinary differential equations in Banach spaces were investigated in many papers (see [–]). In reflexive Banach spaces, the existence of weak solutions or pseudo-solutions for fractional differential equations were studied in [–]. In this section we establish an existence result for the following fractional differential equation:  Dαp y(t) = f (t, y(t)), y() = y ,

()

where Dαp y(·) is a fractional pseudo-derivative of the function y(·) : T → E and f (·, ·) : T × E → E is a given function. Along with the Cauchy problem () consider the following integral equation: 

t

y(t) = y + 

 (t – s)α–  f s, y(s) ds, (α)

t ∈ T,

()

where the integral is in the sense of Pettis. A continuous function y(·) : T → E is said to be a solution of () if y(·) has a pseudoderivative belonging to P∞ (T, E), Dαp y(t)  f (t, y(t)) for t ∈ T and y() = y . To prove a result on the existence of solutions for () we need some preliminary results. Lemma  Let f (·, ·) : T × E → E be a function such that f (·, y(·)) ∈ P∞ (T, E), for every continuous function y(·) : T → E. Then a continuous function y(·) : T → E is a solution of () if and only if it satisfies the integral equation (). Proof Indeed, if a continuous function y(·) : T → E is a solution of (), then y(·) has a pseudo-derivative belonging to P∞ (T, E), Dαp y(t)  f (t, y(t)) for t ∈ T and y() = y . Then we have I α Dαp y(t) = I α f (t, y(t)) on T, and thus from Lemma (b) it follows that y(t) – y() = I α f (t, y(t)) on T; that is, y(·) satisfies the integral equation (). Conversely, suppose that a continuous function y(·) : T → E satisfies the integral equation (). Then the function z(·) := f (·, y(·)) ∈ P∞ (T, E) satisfies the Abel equation  

t

(t – s)α– z(s) ds = v(t), (α)

t ∈ T,

where v(t) := y(t) – y , t ∈ T. Then from [], Theorem ., and Lemma  it follows that v–α (·) has a pseudo-derivative on T and z(t) 

dp dp t –α v–α (t) = y–α (t) – y() for t ∈ T. dt dt ( – α)

Then by Remark  we have z(t)  Dαp y(t) for t ∈ T; that is, Dαp y(t)  f (t, y(t)) on T.



In this section we shall discuss the existence of solutions of fractional differential equations in nonreflexive Banach spaces. We recall that a function f (·) : E → E is said to be sequentially continuous from Ew into Ew (or weakly-weakly sequentially continuous) if, for every weakly convergent sequence {xn }n≥ ⊂ E, the sequence {f (xn )}n≥ is weakly convergent in E.

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 8 of 18

By a Gripenberg function we mean a function g : R+ → R+ such that g(·) is continuous, nonincreasing with g() = , and u ≡  is the only continuous solution of u(t) ≤

 (α)



t

  (t – s)α– g u(s) ds,

u() = .

()



The problem of uniqueness of the null solution of () was studied by Gripenberg in []. Let us denote by Pwk (E) the set of all weakly compact subset of E. The weak measure of noncompactness [] is the set function β : Pwk (E) → [, ∞) defined by   β(A) = inf r > ; there exist K ∈ Pwk (E) such A ⊂ K + rB , where B is the closed unit ball in E. The properties of the weak noncompactness measure are analogous to the properties of the measure of noncompactness, namely (see []): () A ⊆ B implies that β(A) ≤ β(B); () β(A) = β(clw A), where clw A denotes the weak closure of A; () β(A) =  if and only if clw A is weakly compact; () β(A ∪ B) = max{β(A), β(B)}; () β(A) = β(conv(A)); () β(A + B) ≤ β(A) + β(B); () β(x + A) = β(A), for all x ∈ E; () β(λA) = |λ|β(A), for all λ ∈ R;  () β( ≤r≤r rA) = r β(A); () β(A) ≤  diam(A). Lemma  ([]) Let H ⊂ C(T, E) be bounded and equicontinuous. Then (i) the function t → β(H(t)) is continuous on T, (ii) βc (H) = supt∈T β(H(t)), where βc (·) denote the weak noncompactness measure on C(T, E) and H(t) = {u(t), u ∈ H}, t ∈ T. Lemma  ([]) Let E be a metrizable locally convex topological vector space and let K be a closed convex subset of E, and let Q be a weakly sequentially continuous map of K into itself. If for some y ∈ K the implication   V = conv Q(V ) ∪ {y}

⇒

V is relatively weakly compact

holds, for every subset V of K , then Q has a fixed point. Theorem  Assume f (·, ·) : T × E → E is a function such that: (H) f (t, ·) is weakly-weakly sequentially continuous, for every t ∈ T; (H) f (·, y(·)) ∈ P∞ (T, E), for every continuous function y(·) : T → E; (H) f (t, y) ≤ M, for all (t, y) ∈ T × E; (H) for every bounded set A ⊆ E we have     β f (T × A) ≤ g β(A) ,

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 9 of 18

where g(·) is a Gripenberg function. Then () admits a solution y(·) on an interval T = [, a] with a = min{b, ( (α+) )/α }. M Proof In our proof we shall use some ideas from [] and []. We define the nonlinear operator Q(·) : C(T , E) → C(T , E) by  (Qy)(t) = y + 

t

 (t – s)α–  f s, y(s) ds, (α)

t ∈ T .

If y(·) ∈ C(T , E), then by (H) we have f (·, y(·)) ∈ P∞ (T, E) and so the operator Q makes sense. To show that Q is well defined, let t , t ∈ T with t > t . Without loss of generality, assume that (Qy)(t ) – (Qy)(t ) = . Then by the Hahn-Banach theorem, there exists a y∗ ∈ E∗ with y∗  =  and (Qy)(t ) – (Qy)(t ) = |y∗ , (Qy)(t ) – (Qy)(t )|. Then   (Qy)(t ) – (Qy)(t )   =  y∗ , (Qy)(t ) – (Qy)(t )     t  t     (t – s)α–  ∗  (t – s)α–  ∗   y , f s, y(s) ds – y , f s, y(s) ds = (α) (α)     t    (t – s)α– (t – s)α–  ∗  – y , f s, y(s)  ds ≤ (α) (α)   t   (t – s)α–  ∗  y , f s, y(s)  ds + (α) t ≤

 M M α α t – t + (t – t )α ≤ (t – t )α , ( + α)  ( + α)

()

so Q maps C(T , E) into itself. Let K be the convex, closed, and equicontinuous set defined by      K = y(·) ∈ C(T , E); y(·)c ≤ y  + , y(t ) – y(t ) ≤

 M |t – t |α , for all t , t ∈ T . ( + α)

We will show that Q maps K into itself and Q restricted to the set K is weakly-weakly sequentially continuous. To show that Q : K → K , let y(·) ∈ K and t ∈ T . Again, without loss of generality, assume that (Qy)(t) = . By the Hahn-Banach theorem, there exists a y∗ ∈ E∗ with y∗  =  and (Qy)(t) = |y∗ , (Qy)(t)|. Then by (H), we have   (Qy)(t) ≤ y  +

 

≤ y  +

t

  (t – s)α–  ∗  y , f s, y(s)  ds (α)

Maα ≤ y  + , (α + )

and using () it follows that Q maps K into K . Next, we show that Q is weakly-weakly sequentially continuous. First, we recall that the weak convergence in K ⊂ C(T , E) is exactly the weak pointwise convergence. Let {yn (·)}n≥ be a sequence in K such that yn (·) converges weakly to y(·) in K . Then yn (t) converges weakly to y(t) in E for each t ∈ T .

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 10 of 18

Since K is a closed convex set, by Mazur’s lemma we have y(·) ∈ K . Further, by (H) it follows that f (t, yn (t)) converges weakly to f (t, y(t)) for each t ∈ T . Then the Lebesgue dominated convergence theorem for the Pettis integral (see []) yields I α yn (t) converging weakly to I α y(t) in E for each t ∈ T . Since K is an equicontinuous subset of C(T , E) it follows that Q(·) is weakly-weakly sequentially continuous. Suppose that V ⊂ K is such that V = co(Q(V ) ∪ {y(·)}) for some y(·) ∈ K . We will show that V is relatively weakly compact in C(T , E). Let 

t



 (t – s)α–  f s, V (s) ds = (α)

and (QV )(t) = y +

 

t

 (t – s)α–  f s, y(s) ds; y(·) ∈ V (α)

t

( ε(α+) )/α and M ∗ ∗

(t–s)α– f (s, V (s)) ds. Let t (α) t (t–s)α– t–η (α) f (s, y(s)) ds =  ∗ 

η< exists a y ∈ E with y  =  and    

t t–η



∈ T and ε > . If we choose η >  such that then, by the Hahn-Banach theorem, there

       ∗ t (t – s)α–    (t – s)α–    f s, y(s) ds =  y , f s, y(s) ds . (α) (α) t–η

It follows that    

t t–η

  t     (t – s)α–  (t – s)α–  ∗  f s, y(s) ds ≤ y , f s, y(s)  ds ≤ ε,  (α) (α) t–η

and thus using property () of the noncompactness measure we infer 

t

β t–η

  (t – s)α–  f s, V (s) ds ≤ ε. (α)

()

Since by Lemma  the function s → v(s) := β(V (s)) is continuous on [, t – η] it follows that s → (t – s)α– g(v(s)) is continuous on [, t – η]. Hence, there exists δ >  such that      (t – τ )α– g v(τ ) – (t – s)α– g v(s)  < ε  and      g v(ξ ) – g v(τ )  <

ε , ηα–

for all τ , s, ξ ∈ [, t – η] with |τ – s| < δ and |τ – ξ | < δ. It follows that      (t – τ )α– g v(ξ ) – (t – s)α– g v(s)            ≤ (t – τ )α– g v(τ ) – (t – s)α– g v(s)  + (t – τ )α– g v(ξ ) – g v(τ )  < ε, that is,      (t – τ )α– g v(ξ ) – (t – s)α– g v(s)  < ε,

()

for all τ , s, ξ ∈ [, t – ξ ] with |τ – s| < δ and |τ – ξ | < δ. Consider a partition of the interval [, t – η] into n parts  = t < t < · · · < tn = t – η such that ti – ti– < δ, i = , , . . . , n.

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 11 of 18

From Lemma  it follows that for each i ∈ {, , . . . , n} there exists si ∈ [ti– , ti ] such that β(V ([ti– , ti ])) = v(si ), i = , , . . . , n. Then we have (see [], Theorem .) 

 (t – s)α–  f s, V (s) ds (α)  n      ti (t – s)α– f s, V (s) ds ⊂ (α) i= ti– t–n

      (ti – ti– )conv (t – s)α– f s, y(s) ; s ∈ [ti– , ti ], y ∈ V , (α) i= n

⊂

and so 

t–η

β 

 (t – s)α–  f s, V (s) ds (α)



       (ti – ti– )β conv (t – s)α– f s, y(s) ; s ∈ [ti– , ti ], y ∈ V (α) i= n

≤

      = (ti – ti– )β (t – s)α– f s, y(s) ; s ∈ [ti– , ti ], y ∈ V (α) i= n

≤

     (ti – ti– )(t – ti )α– β f T × V [ti– , ti ] (α) i=

≤

     (ti – ti– )(t – ti )α– g β V [ti– , ti ] (α) i=

=

    (ti – ti– )(t – ti )α– g v(si ) . (α) i=

n

n

n

Using () we have      (t – ti )α– g v(si ) – (t – s)α– g v(s)  < ε(α + ). This implies that     (ti – ti– )(t – ti )α– g v(si ) ≤ (α) i= n



t–η



  (t – s)α–   g v(s) ds + ε t α – ηα . (α)

Thus we obtain 

t–η

β 

  t–η    (t – s)α–  (t – s)α–   f s, V (s) ds ≤ g v(s) ds + ε t α – ηα . (α) (α) 

Now because  (QV )(t) ⊂ 

t–η

 (t – s)α–  f s, V (s) ds + (α)



t t–η

 (t – s)α–  f s, V (s) ds, (α)

()

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 12 of 18

then from () and () we have 





t–η

β (QV )(t) ≤ 



t

≤ 

  (t – s)α–   g v(s) ds + ε t α – ηα + ε (α)

  (t – s)α–   g v(s) ds + ε t α – ηα +  . (α)

As the last inequality is true, for every ε > , we infer 





β (QV )(t) ≤

t



(t – s)α–   g v(s) ds. (α)

Because V = co(Q(V ) ∪ {y(·)}) then         β V (t) = β co Q(V ) ∪ y(·) ≤ β (QV )(t) and thus  v(t) ≤ 

t

(t – s)α–   g v(s) ds for t ∈ T . (α)

Since g(·) is a Gripenberg function, it follows that v(t) =  for t ∈ T . Since V as a subset of K is equicontinuous, by Lemma  we infer     βc V (T ) = sup β V (t) = . t∈T

Thus, by Arzelá-Ascoli’s theorem we find that V is weakly relatively compact in C(T , E). Using Lemma  there exists a fixed point of the operator Q which is a solution of ().  If E is reflexive and f (·, ·) : T × E → E is bounded, then (H) is automatically satisfied since a subset of a reflexive Banach space is weakly compact iff it is closed in the weak topology and bounded in the norm topology. If for α =  we put Dp y(·) = y p (·), then from Theorem  we obtain the following result (see [, ]). Corollary  If f (·, ·) : T × E → E is a function that satisfies the conditions (H)-(H) in Theorem , then the differential equation  y p (t) = f (t, y(t)), y() = y ,

()

has a solution on [, a] with a = min{b, /M}.

4 Multi-term fractional differential equation The case of multi-term fractional differential equations in reflexive Banach spaces was recently considered in [–]. Consider the following multi-term fractional differential

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 13 of 18

equation:  D

αm

–

m– 

 ai D

αi

  y(t) = f t, y(t) for t ∈ [, ],

y() = ,

()

i=

where Dαi y(·), i = , , . . . , m, are fractional pseudo-derivatives of order αi ∈ (, ) of a pseudo-differentiable function y(·) : [, ] → E, f (t, ·) : [, ] × E → E is weakly-weakly sequentially continuous, for every t ∈ [, ], and f (·, y(·)) is Pettis integrable, for every continuous function y(·) : [, ] → E, E is a nonreflexive Banach space,  < α < α < · · · < αm <   |ai | and a , a , . . . , am– are real numbers such that a := m– i= (αm –αi +) < . Along with the Cauchy problem () consider the following integral equation:

y(t) =

m– 

  ai I αm –αi y(t) + I αm f t, y(t) ,

()

i=

t ∈ T, where the integral is in the sense of Pettis and T = [, ]. A continuous function y(·) : T → E is said to be a solution of () if (i) y(·) has Caputo fractional pseudo-derivatives of orders αi ∈ (, ), i = , , . . . , m,  αi (ii) (Dαm – m– i= ai D )y(t)  f (t, y(t)), for all t ∈ T, (iii) y() = . Lemma  Assume that f (·, ·) : T × E → E satisfy the assumptions (H) and (H) in Theorem . Then every continuous function y(·) : T → E which satisfies the integral equation () is a solution of (). Proof Suppose that a continuous function y(·) : T → E satisfies the integral equation (). Then z(·) := f (·, y(·)) ∈ P∞ (T, E) satisfies the Abel equation 

t



(t – s)αm – z(s) ds = v(t), (αm )

t ∈ T,

 αm –αi where v(t) := y(t) – m– y(t), t ∈ T. From [], Theorem ., it follows that v–αm (·) i= ai I has a pseudo-derivative on T and z(t) 

dp v–αm (t) for t ∈ T, dt

Since y(·) is continuous on T and f (·, y(·)) ∈ P∞ (T, E) satisfies (H), we have   lim+ I α y(t) = lim+ I α f t, y(t) = 

t→

t→

for α ∈ (, ) and thus, taking the limit as t → + on both equalities in (), we obtain y() =  and consequently v() = . Since v() = , by Remark  we have z(t) 

dp v–αm (t) = Dpαm v(t) = Dαp m v(t), dt

t ∈ T.

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 14 of 18

Since by Lemma (b) we have

Dαp m v(t) = Dαp m y(t) –

m– 

ai Dαp m I αm –αi y(t) = Dαp m y(t) –

i=

m– 

ai Dαp i y(t),

i=

we obtain  Dαp m –

m– 



  ai Dαp i y(t)  f t, y(t) ,

t ∈ T.

i=

Hence the continuous function y(·) satisfy the conditions (i)-(iii) from definition and thus y(·) is a solution of ().  Lemma  ([], Theorem .) Let K be a nonempty, bounded, convex, closed set in a Banach space E. Assume Q : K → K is weakly sequentially continuous and β-contractive (that is, there exists  ≤ k <  such that β(Q(A)) ≤ k β(A), for all bounded sets A ⊂ E). Then Q has a fixed point. Remark  Since the function σ → (σ ) is convex and (σ ) ≥ (/) ≈ . for σ ∈ (, ), for every r ∈ (, (/)) we have (αm + ) > r. Next we establish an existence result for the multi-term fractional integral equation () in nonreflexive Banach spaces. Theorem  Suppose that f (·, ·) : T ×E → E satisfies the conditions (H)-(H) in Theorem  and there exists L >  such that, for every bounded set A ⊆ E, we have   β f (T × A) ≤ Lβ(A). If r ∈ (, ) is such that (αm + ) > r, then () admits a solution y(·) on an interval T = [, a ] with     ( – a)(αm + ) /αm r . a < min , r+L M Proof We define the nonlinear operator Q(·) : C(T , E) → C(T , E) by

(Qy)(t) =

m– 

  ai I αm –αi y(t) + I αm f t, y(t) ,

i=

for all t ∈ T . We remark that a solution of integral equation () is a fixed point of the operator Q. If y(·) ∈ C(T , E), then by (H) we have f (·, y(·)) ∈ P∞ (T , E) and so the operator Q makes sense. To show that Q is well defined, let t, s ∈ T with t > s. Without loss of generality, assume that (Qy)(t) – (Qy)(s) = . Then by the Hahn-Banach theorem, there

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 15 of 18

exists a y∗ ∈ E∗ with y∗  =  and (Qy)(t) – (Qy)(s) = |y∗ , (Qy)(t) – (Qy)(s)|. Then   (Qy)(t) – (Qy)(s)   =  y∗ , (Qy)(t) – (Qy)(s)  ≤

m–  i=



|ai | (αm – αi ) t

(t – τ )

+



s

  (s – τ )αm –αi – – (t – τ )αm –αi –  y∗ , y(τ )  dτ



 ∗  y , y(τ )  dτ

αm –αi – 



s

 s       (s – τ )αm – – (t – τ )αm –  y∗ , f τ , y(τ )  dτ (αm )    t     + (t – τ )αm –  y∗ , f τ , y(τ )  dτ

+

s

≤

 m–  i=

 M |ai | yc + (t – s)αm , (αm – αi + ) (αm + )

()

so Q maps C(T , E) into itself. Let δ ≥  and let K be the convex, closed, bounded and equicontinuous set defined by 

    K = y(·) ∈ C(T , E); y(·)c ≤ δ, y(t) – y(s) ≤

 m–  i=

  M δ|ai | αm + |t – s| , for all t, s ∈ T . (αm – αi + ) (αm + )

Without loss of generality, assume that (Qy)(t) = . By the Hahn-Banach theorem, there exists a y∗ ∈ E∗ with y∗  =  and (Qy)(t) = |y∗ , (Qy)(t)|. Then by (H), we have     (Qy)(t) =  y∗ , (Qy)(t)  ≤

m–  i=

≤

m–  i=



t

|ai | 

 (t – s)αm –αi –  ∗ y , y(τ )  ds + (αm – αi )



t 

  (t – s)αm –  ∗  y , f τ , y(τ )  ds (αm )

Maαm δ|ai | + ≤ δa + ( – a)δ = δ (αm – αi + ) (αm + )

and using () it follows that Q maps K into K . Following the same reasoning as in the proof of Theorem  it is easy to show that Q is weakly-weakly sequentially continuous from K to K . Next, we will prove that Q has at least one fixed point y (·) ∈ K . Let V ⊂ K be such that βc (V ) > . Next, to simplify the writing of some relations, we will use the following notations: 

 (t – s)αm –  f s, y(s) ds, (αm )   m– t–η  (t – s)αm –αi – y(s) ds, B(t) := ai (αm – αi )  i=

A(t) :=

t–η

Agarwal et al. Advances in Difference Equations (2015) 2015:112



Page 16 of 18

 (t – s)αm –  f s, y(s) ds, (αm ) t–η  m– t  (t – s)αm –αi – D(t) := ai y(s) ds, t–η (αm – αi ) i= t

C(t) :=

 r|ai |ηαm –αi Mηαm and |y∗ , D(t)| ≤ m– for t ∈ T . Then it is easy to see that |y∗ , C(t)| ≤ (α i= (αm –αi +) , m +) for all y∗ ∈ E∗ with y∗  = . Let t ∈ T and ε > . If we choose η >  such that η < ε(αm +) ( r[M+(α )/αm and C(t) + D(t) = , then by the Hahn-Banach theorem, there exists a m +)] ∗ ∗ y ∈ E with y∗  =  and     C(t) + D(t) =  y∗ , C(t) + D(t)  ≤

m–  Mηαm r|ai |ηαm –αi + (αm – αi + ) (αm + ) i=

≤ rηαm +

M + (αm + ) αm Mηαm ≤r η < ε, (αm + ) (αm + )

and thus using property () of the measure of noncompactness we infer   β (CV )(t) + (DV )(t) ≤ ε. As in the proof of Theorem , with g(t) = L, t ∈ T , we obtain 





t–η

β (AV )(t) ≤ L 

   (t – s)αm –  β V (s) ds + ε t α – ηα . (αm )

Also, with y(·) instead of f (·, y(·)), we have   β ai

t–η





(t – s)αm –αi – V (s) ds (αm – αi ) t–η

≤ |ai | 



  ε α (t – s)αm –αi –  β V (s) ds + t – ηα , (αm – αi ) m–

and so 



β (BV )(t) ≤

m– 



t–η

|ai | 

i=

   (t – s)αm –αi –  β V (s) ds + ε t α – ηα . (αm – αi )

Next, since (QV )(t) = (AV )(t) + (CV )(t) + (BV )(t) + (DV )(t), t ∈ T , then from the last inequalities and using properties of the noncompactness measure we infer         β (QV )(t) ≤ β (AV )(t) + β (BV )(t) + β (CV )(t) + (DV )(t) 

  (t – s)αm –  |ai | β V (s) ds + (αm )  i=   + ε t α – ηα + ε.

≤L

t

m–



t 

 (t – s)αm –αi –  β V (s) ds (αm – αi )

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 17 of 18

As the last inequality is true, for every ε > , it follows that m–        β (QV )(t) ≤ LI αm β V (t) + |ai |I αm –αi β V (t) ,

t ∈ T .

i=

Since β(V (t)) ≤ βc (V ), t ∈ T , we have   β (QV )(t) ≤



   a L Lt αm + t αm –αi βc (V ) ≤ + a βc (V ) ≤ k βc (V ), (αm + ) r

where k = a ( + Lr ) < . It follows that βc (QV ) < k βc (V ), for every set V ⊂ K with βc (V ) > ; that is, Q : K → K is a βc -contractive operator. Since K is a nonempty, closed, convex, bounded subset in C(T , E), and Q : K → K is weakly sequentially continuous and βc -contractive, by Lemma  it follows that the operator Q has a fixed point y (·) ∈ K .  Using Lemma  we obtain the following result. Corollary  If the assumptions of Theorem  are satisfied, then the problem () has at least one solution.

Competing interests The authors declare that they have no competing interests. Authors’ contributions All authors contributed equally to the writing of this paper. All authors read and approved the final manuscript. Author details 1 Department of Mathematics, Texas A&M University-Kingsville, Kingsville, TX 78363, USA. 2 Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah, 21589, Saudi Arabia. 3 Constantin Brancusi University, Republicii 1, Targu-Jiu, 210152, Romania. 4 School of Mathematics, Statistics and Applied Mathematics, National University of Ireland, Galway, Ireland. 5 Department of Mathematics and Statistics, University of Swat, Khyber Paukhtunkhwa, Pakistan. Acknowledgements The authors wish to thank Professor Mieczysław Cichon´ for his helpful comments. Received: 12 January 2015 Accepted: 25 March 2015 References 1. Agarwal, RP, Lupulescu, V, O’Regan, D, Rahman, G: Fractional calculus and fractional differential equations in nonreflexive Banach spaces. Commun. Nonlinear Sci. Numer. Simul. 20(1), 59-73 (2015) 2. Agarwal, RP, Lupulescu, V, O’Regan, D, Rahman, G: Multi-term fractional differential equations in a nonreflexive Banach space. Adv. Differ. Equ. 2013, 302 (2013). doi:10.1186/1687-1847-2013-302 3. Agarwal, RP, Lupulescu, V, O’Regan, D, Rahman, G: Weak solution for fractional differential equations in nonreflexive Banach spaces via Riemann-Pettis integrals. Math. Nachr. (in press) 4. Ben Amar, A: Some fixed point theorems and existence of weak solutions of Volterra integral equation under Henstock-Kurzweil-Pettis integrability. Comment. Math. Univ. Carol. 52, 177-190 (2011) ´ M: On bounded weak solutions of a nonlinear differential equation in Banach spaces. Funct. Approx. 5. Cichon, Comment. Math. 21, 27-35 (1992) 6. Salem, HAH: On the nonlinear Hammerstein integral equations in Banach spaces and application to the boundary value problem of fractional order. Math. Comput. Model. 48, 1178-1190 (2008) ´ M: On solutions of fractional order boundary value problems with integral boundary conditions 7. Salem, HAH, Cichon, in Banach spaces. J. Funct. Spaces Appl. 2013, Article ID 428094 (2013) 8. Dinuleanu, N: Vector Measures. Pergamon, New York (1967) 9. Singer, I: Linear functionals on the space of continuous mappings of a compact space into a Banach space. Rev. Roum. Math. Pures Appl. 2, 301-315 (1957) 10. Mitchell, AR, Smith, C: An existence theorem for weak solutions of differential equations in Banach spaces. In: Lakshmikantham, V (ed.) Nonlinear Equations in Abstract Spaces, pp. 387-404 (1978) 11. Solomon, BW: On differentiability of vector-valued functions of a real variables. Stud. Math. 29, 1-4 (1967) 12. Diestel, J, Uhl, JJ Jr.: Vector Measures. Mathematical Surveys, vol. 15. Am. Math. Soc., Providence (1977)

Agarwal et al. Advances in Difference Equations (2015) 2015:112

Page 18 of 18

13. Musial, K: Topics in the Theory of Pettis Integration. In School of Measure Theory and Real Analysis, Grado, Italy, May (1992) 14. Schwabik, S, Guoju, Y: Topics in Banach Space Integration. World Scientific, Singapore (2005) 15. Pettis, JP: On integration in vector spaces. Trans. Am. Math. Soc. 44, 277-304 (1938) 16. Salem, HAH, El-Sayed, SMA, Moustafa, OL: A note on the fractional calculus in Banach spaces. Studia Sci. Math. Hung. 42(2), 115-130 (2005) 17. Arendt, W, Batty, C, Hieber, M, Neubrander, F: Vector-Valued Laplace Transforms and Cauchy Problems. Monogr. Math., vol. 96. Birkhäuser, Basel (2001) 18. Knight, WJ: Solutions of differential equations in Banach spaces. Duke Math. J. 41, 437-442 (1974) 19. Cramer, E, Lakshmikantham, V, Mitchell, AR: On the existence of weak solutions of differential equations in nonreflexive Banach spaces. Nonlinear Anal. 2, 259-262 (1978) 20. Arino, O, Gautier, S, Penot, JP: A fixed point theorem for sequentially continuous mappings with application to ordinary differential equations. Funkc. Ekvacioj 27, 273-279 (1984) 21. Kubiaczyk, I: On a fixed point theorem for weakly sequentially continuous mapping. Discuss. Math., Differ. Incl. 15, 15-20 (1995) ´ M, Kubiaczyk, I: Kneser’s theorem for strong, weak and pseudo-solutions of ordinary differential equations in 22. Cichon, Banach spaces. Ann. Pol. Math. 52, 13-21 (1995) ´ M: Weak solutions of ordinary differential equations in Banach spaces. Discuss. Math., Differ. Incl. Control 23. Cichon, Optim. 15, 5-14 (1995) 24. O’Regan, D: Fixed point theory for weakly sequentially continuous mapping. Math. Comput. Model. 27(5), 1-14 (1998) 25. O’Regan, D: Weak solutions of ordinary differential equations in Banach spaces. Appl. Math. Lett. 12, 101-105 (1999) 26. Szufla, S: Kneser’s theorem for weak solutions of an mth-order ordinary differential equation in Banach spaces. Nonlinear Anal. 38, 785-791 (1999) 27. O’Regan, D: Fixed point theory for weakly sequentially closed maps. Arch. Math. 36, 61-70 (2000) ´ M, Kubiaczyk, I, Sikorska-Nowak, A: The Henstock-Kurzweil-Pettis integrals and existence theorems for the 28. Cichon, Cauchy problem. Czechoslov. Math. J. 54, 279-289 (2004) 29. Dutkiewicz, A, Szufla, S: Kneser’s theorem for weak solution of and integral equation with weakly singular kernel. Publ. Inst. Math. (Belgr.) 77(91), 87-92 (2005) ´ M: On solutions of differential equations in Banach spaces. Nonlinear Anal. 60, 651-667 (2005) 30. Cichon, ´ M, Kubiaczyk, I, Sikorska-Nowak, A, Yantir, A: Weak solutions for the dynamic Cauchy problem in Banach 31. Cichon, spaces. Nonlinear Anal. 71, 2936-2943 (2009) 32. Salem, HAH, El-Sayed, AMA: Weak solution for fractional order integral equations in reflexive Banach space. Math. Slovaca 55(2), 169-181 (2005) 33. Benchohra, M, Graef, JR, Mostefai, F: Weak solutions for nonlinear fractional differential equations on reflexive Banach spaces. Electron. J. Qual. Theory Differ. Equ. 2010, 54 (2010) 34. Salem, HAH: On the fractional calculus in abstract spaces and their applications to the Dirichlet-type problem of fractional orders. Comput. Math. Appl. 59, 1278-1293 (2010) 35. Benchohra, M, Mostefai, F: Weak solutions for nonlinear fractional differential equations with integral boundary conditions in Banach spaces. Opusc. Math. 32(1), 31-40 (2012) 36. Hernández, E, O’Regan, D, Balachandran, K: Existence results for abstract fractional differential equations with nonlocal conditions via resolvent operators. Indag. Math. 24, 68-82 (2013) 37. Gripenberg, C: Unique solutions of some Volterra integral equations. Math. Scand. 48, 59-67 (1981) 38. De Blasi, F: On a property of the unit sphere in a Banach space. Bull. Math. Soc. Sci. Math. Roum. 21, 259-262 (1977) 39. Ambrosetti, A: Un teorema di esistenza per le equazioni differenziali negli spazi di Banach. Rend. Semin. Mat. Univ. Padova 39, 349-369 (1967) 40. Talagrand, M: Pettis Integral and Measure Theory. Mem. Amer. Math. Soc., vol. 307. Am. Math. Soc., Providence (1984) 41. Geitz, RF: Geometry and the Pettis integral. Trans. Am. Math. Soc. 169(2), 535-548 (1982) 42. Daftardar-Gejji, V, Jafari, H: Solving a multi-order fractional differential equation using Adomian decomposition. Appl. Math. Comput. 189, 541-548 (2007) 43. Salem, HAH: Monotonic solutions of multi-term fractional differential equations. Comment. Math. Prace Mat. 47(1), 1-7 (2007) 44. Salem, HAH: Multi-term fractional differential equations in Banach spaces. Math. Comput. Model. 49, 829-834 (2009) 45. Salem, HAH: Global monotonic solutions of multi-term fractional differential equations. Appl. Math. Comput. 217, 6597-6603 (2011)