THE (k, s)-FRACTIONAL CALCULUS OF CLASS OF A FUNCTION
GAUHAR RAHMAN, ABDUL GHAFFAR, K. S. NISAR*, AZEEMA
Abstract. In this present paper, we deal with the generalized (k, s)-fractional integral
and differential operators recently defined by Nisar et al. and obtain some generalized (k, s)-fractional integral and differential formulas involving the class of a function as its
kernels. Also, we investigate a certain number of their consequences containing the said function in their kernels.
1. introduction
Fractional calculus has gained appreciable fame and significance due to its numerous and boundless applications in Science; more particularly in engineering. For the recent development in the field of fractional calculus, one can refer to [5, 13, 15] and [6, 7]. One more direction to such study was proposed by Atangana and Baleanu [1] by introducing derivatives which are based upon the generalized Mittag-Leffler function. Integral inequal-ities are considered to be of highly importance because they are useful in the research of different subjects of differential and integral equations (see [8]). We begin with the work of Diaz and Pariguan [3] which is defined as:
(p)n,k = (
p(p+k), ..., (p+ (n−1)k) (n ∈N, p ∈C),
1, (n= 0, p∈C), (1.1)
and
Γk(η) = Z ∞
0
tη−1e−tkk dt, (1.2)
where η∈C, k >0, <(z)>0.
They also define the following relations:
Γk(η+k) = ηΓk(η), (1.3)
2010Mathematics Subject Classification. 26B25, 33C45, 33B15, 26A33.
Key words and phrases. fractional integral operators; fractional differential operators; generalized (k, s)-fractional integral operators; generalized (k, s)-fractional differential operators.
*Corresponding author
G. Rahman: Department of Mathematics, International Islamic University, Islamabad, Pakistan;
A. Ghaffar: Department of Mathematical Science, BUITEMS Quetta, Pakistan;
K. S. Nisar: Department of Mathematics, College of Arts and Science, Prince Sattam bin Abdulaziz
Uni-versity, Wadi Al dawaser, Riyadh region 11991, Saudi Arabia; [email protected], [email protected] Azeema: Department of Mathematics, SBK Women’s University, Quetta, Pakistan;
1
and
Γk(η) =k η k−1Γ
η
k
. (1.4)
A new generalization of the k-fractional integral has been defined by Mubeen and Habibul-lah [11] as follows:
Ikµ(f(x)) = 1 kΓk(µ)
Z x
0
(x−τ)µk−1f(τ)dτ. (1.5)
Clearly, it is observed that whenk = 1 (1.5) coincide with the result of Riemann-Liouville (R-L) fractional integration formula (see [5]). In fact, the following particular cases:
Iµ(f(x)) = 1 Γ(µ)
Z x
0
(x−τ)µ−1f(τ)dτ. (1.6)
In the same paper, they also define the following results:
Ikρ(xβk−1) = Γk(β)
Γk(ρ+β)
xkρ+ β
k−1, (1.7)
and
Ikρ((x−u)βk−1) = Γk(β)
Γk(ρ+β)
(x−u)ρk+ β
k−1. (1.8)
Sarikaya et al. [14] have developed the R-L (k, s)-fractional integral of order µ > 0 is defined by:
s kI
µ
af(x) =
(s+ 1)1−µk
kΓk(µ) Z x
0
(xs+1−ts+1)µk−1tsf(t)dt, (1.9)
where s∈ < \ {−1}, k > 0 andx∈[a;b]. Also, they define,
s kI
µ a[(x
s+1−as+1)λk−1] = Γk(λ)
(s+ 1)µkΓk(λ+µ)
(xs+1−as+1)λ+kµ−1. (1.10)
For more details about (k, s)-fractional integrals interesting readers can refer to [2, 12, 16]. Recently Nisar et al., [9, 10] defined the following R-L left and right sided (k, s)-fractional integral and differential operators of order µas:
s a,kD
−µ
x f(x) = s a,kI
µ
xf(x) = s kI
µ a+f(x)
= (skIaµ+f)(x) = (s+ 1) 1−µk
kΓk(µ) Z x
a
f(t)
(xs+1−ts+1)1−µkt
sdt,(x > a),
(1.11)
s ρ,kD
−µ
β f(x) = s ρ,kI
µ
βf(x) = s k I
µ a−f(x)
= (skIaµ−f)(x) = (s+ 1) 1−µ
k
kΓk(µ) Z β
x
f(t) (xs+1−ts+1)1−µ
k
tsdt,(x < β),
(1.12)
(skDaµ+f)(x) = 1 xs
d dx
n
and
(skDaµ−f)(x) =− 1
xs
d dx
n
kn skIank−−µf(x), (1.14) respectively. Substituting k = 1 and s = 0, then the above relation will coincide to the R-L left and right sided (k, s)-fractional integral and derivatives see ([5],[7]).
The R-L left and right sided (k, s)-fractional derivative operators kD
µ
a+ defined in (1.12) is generalized by (k, s)-fractional derivative operator is denoted by s
kD µ,ν
a+, whereµ is the order such that 0 < ν < 1, we define as
(skDaµ,ν+f)(x) =
s kI
ν(k−µ)
ρ+
1 xs
d dx
(kn skIa(1+−ν)(k−µ)f)(x), (1.15)
Obviously, when v = 0 then (1.15) approaches to the R-L (k, s)-fractional derivatives operator s
kD µ
a+ (1.13).
They [9] also defined the following lemma as:
Lemma 1.1. For k > 0, with x > ρ, 0 < ν < 1 and <(λ) > 0, then the following result for (k, s)-fractional derivatives operator skDaµ+ in (1.15) hold true:
(skDaµ,ν+(τs+1−as+1)λk−1)(x) = Γk(λ)
(s+ 1)−µkΓk(λ−µ)
(xs+1−as+1)λ−kµ−1. (1.16)
For the present investigation, we consider the following class of a function recently defined by Tunc [17]
zσ,kρ,λ(x) =
∞ X
n=0
σ(m) kΓk(ρkm+λ)
xm,(ρ, λ >0;|x|<R), (1.17)
where the coefficients σ(m)(m ∈ N0 = N ∪ {0} is a bounded sequence of positive real numbers R is the set of real numbers.
2. Generalized (k, s)-fractional integrals and differentials formulas
In this section, we present the generalized (k, s)-fractional integrals and differentials formulas involving a class of a function zσ,kρ,λ(x) as defined in (1.17). In this continuation of the study of generalizedk-fractional calculus, we define the following fractional integral operator.
Definition 2.1. If k >0, ρ, δ, ω ∈ C; with R(ρ)>0, R(β)> 0, R(δ)> 0, and x > ρ, then
(skεωa+;;σρ,λf)(x) = 1 k
Z x
0
(xs+1−τs+1)λk−1zσ,k ρ,λ(w(x
s+1−τs+1)ρ)τsf(τ)dτ, (2.1)
By putting s= 0, then (2.1) can be written as
(kεωa+;;σρ,λf)(x) =
1 k
Z x
0
(x−τ)λk−1zσ,k ρ,λ(w(x
see [17]. Similarly, when ω = 0 andk = 1 then (2.2) turns to:
(Iaλ+f)(x) = 1 Γ(λ)
Z x
0
f(τ)
(x−τ)1−λdτ,(R(µ)>0). (2.3)
To prove the generalized (k, s)-fractional integral and differential formulas of a class of a function, we first prove the following result.
Lemma 2.1. For k >0, the following result holds true:
1 xs
d dx
m n
(τs+1−as+1)λk−1zα,k ρ,λ[w(x
s+1−
as+1)]ρ
o
=k−m(s+ 1)m(xs+1−as+1)λk−m−1zα,k
ρ,λ−mk[w(x
s+1−as+1)]ρ (2.4) where s ∈ R\ {−1}; ω, α, λ, ρ∈C; R(ω)>0, R(λ)>0, R(ρ)>0 and R(α)>0. Proof. LetS1 be the L. H. S. of (2.4) then
S1 =
1 xs
d dx
m n
(τs+1−as+1)λk−1zα,k ρ,λ[w(x
s+1−as+1)]ρo
=
1 xs
d dx
m n
(τs+1−as+1)λk−1
∞ X
n=0
σ(n) kΓk(ρkn+λ)
[w(xs+1−as+1)ρ]no.
Changing the order of summation and differentiation, we have
S1 =
∞ X
n=0
σ(n) kΓk(ρkn+λ)
[w]n
1 xs
d dx
m
(τs+1−as+1)λk+ρn−1.
(2.5) Now we find
1 xs
d dx
m
(xs+1−as+1)λk+ρn−1
= (s+ 1)m(ρn+λ
k −1)...(ρn+ λ
k −m)(x
s+1−as+1)λk+ρn−m−1
= (s+ 1)m Γ(ρn+
λ k)
Γ(ρn+ λk −m)(x
s+1−as+1)λk+ρn−m−1 (2.6)
Using (1.4), we have
Γ(ρn+λk) Γ(ρn+ λk −m) =
Γk(nρk+λ)
kmΓ
k(nρk+λ−mk)
. (2.7)
Using (2.7) in (2.6), we get
1 xs
d dx
m
(xs+1−as+1)λk+ρn−1
= (s+ 1)m Γk(nρk+λ) kmΓ
k(nρk+λ−mk)
(xs+1−as+1)kλ+ρn−m−1. (2.8)
Thus by using (2.8) in (2.5), we get the following required result. S1 = k−m(s+ 1)m(xs+1−as+1)
λ
k−m−1zα,k
ρ,λ−mk[w(x
Theorem 2.1. For k >0, the following integral formulas holds true: s
kI µ
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)
= (x
s+1−as+1)λ+kµ−1 (s+ 1)µk z
α,k
ρ,λ+µ[w(x
s+1−as+1)]ρ.
(2.10)
where x > a(a ∈ R+ = [0,∞)); ω, α, λ, ρ ∈ C; R(ω) > 0, R(λ) > 0, R(ρ) > 0, and
R(µ)>0.
Proof. LetS2 be L. H. S. side of (2.10) then S2 = skI
µ
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)
= (s+ 1) 1−µ
k
kΓk(µ) Z x
a
(τs+1−as+1)λk−1
zα,kρ,λ[w(τs+1−as+1)ρ](x)
(xs+1−τs+1)1−µ k
dτ
= (s+ 1) 1−µ
k
kΓk(µ) Z x
a
(τs+1−as+1)λk−1
(xs+1−τs+1)1−µk ∞ X
n=0
σ(n) kΓk(nρk+λ)
[w(τs+1−as+1)ρ]ndτ
Changing the order of summation and integration, we have
S2 =
(s+ 1)1−µk
kΓk(µ)
∞ X
n=0
σ(n) kΓk(nρk+λ)
[w]n
Z x
a
(τs+1−as+1)λk+ρn−1
(xs+1−τs+1)1−µk dτ
(2.11) Substitutingτs+1 =as+1+y(xs+1−as+1) in (2.11), then this implies τsdτ = (xs+1s+1−as+1)dy, when τ →a, ⇒ y→0 and τ →x, ⇒ y →1, we have
S2 =
∞ X
n=0
σ(n)[w]n
kΓk(nρk+λ)
(s+ 1)1−µk
kΓk(µ)
×
Z 1
0
(αs+1+y(xs+1−αs+1)−as+1)λk+ρn−1 (xs+1−αs+1−y(xs+1−αs+1))1−µ
k
(xs+1−αs+1) s+ 1 dy
=
∞ X
n=0
σ(n)[w]n
kΓk(nρk+λ)
(xs+1−αs+1)µ+λ+kρnk−1 (s+ 1)µkkΓk(µ)
Z 1
0
yλk+ρn−1(1−y) µ k−1dy
=
∞ X
n=0
σ(n)[w]n kΓk(nρk+λ)
(xs+1−αs+1)µ+λ+kρnk−1
(s+ 1)µkkΓk(µ)
Γk(λ+ρnk)Γk(µ)
Γk(λ+ρnk+µ)
= (x
s+1−αs+1)µ+kλ−1 (s+ 1)µk
∞ X
n=0
σ(n)
kΓk(nρk+λ+µ)
[w(τs+1−as+1)ρ]n
= (x
s+1−as+1)λ+kµ−1 (s+ 1)µk z
α,k
ρ,λ+µ[w(x
s+1−as+1)]ρ.
Theorem 2.2. For k >0, the following result holds true: s
kD µ a+(τ
s+1−
as+1)λk−1)zα,k ρ,λ[w(τ
s+1−
as+1)ρ](x)
= (x
s+1−as+1)λ−kµ−1 (s+ 1)−µk z
α,k
ρ,λ−µ[w(x
s+1−as+1)]ρ.
(2.12)
where x > a(a ∈ R+ = [0,∞)); ω, α, λ, ρ ∈ C; R(ω) > 0, R(λ) > 0, R(ρ) > 0, and
R(µ)>0.
Proof. LetS3 be L. H. S. of (2.12) then
S3 =
1 xs
d dx
n n
kn skIank+−µ(τs+1−as+1)λk−1zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)o
Applying (2.10), this takes the following form
s kD
µ a+(τ
s+1−as+1)λk−1)
zα,kρ,λ[w(τ
s+1−as+1)ρ](x)
=kn
1 xs
d dx
n
(xs+1−as+1)λ−kµ+n−1 (s+ 1)−µk+n z
α,k
ρ,λ−µ+nk[w(x
s+1−as+1)]ρ.
Applying Lemma 2.4, we have
S3 =
(xs+1−as+1)λ−kµ−1
(s+ 1)−µk z α,k
ρ,λ−µ[w(x s+1−
as+1)]ρ.
Theorem 2.3. For k >0, the following result holds true: s
kD µ,ν
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)
= (x
s+1−as+1)λ−kµ−1 (s+ 1)−kµ
zα,kρ,λ−µ[w(x s+1−
as+1)]ρ.
(2.13)
where x > a(a ∈ R+ = [0,∞)); ω, α, λ, ρ ∈ C; R(ω) > 0, R(λ) > 0, R(ρ) > 0, and
R(µ)>0. Proof.
s kD
µ,ν
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)
=sk Dµ,νa+
" ∞ X
n=0
σ(n) kΓk(nρk+λ)
[w(τs+1−as+1)λnk−1+ρ]n #
.
This can be written as
s kD
µ,ν
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
=
∞ X
n=0
σ(n)[w]n
kΓk(nρk+λ) h
s kD
µ,ν
a+(τs+1−as+1)
λ k+ρn−1
i
.
By applying (1.16), we get
s kD
µ,ν
a+(τs+1−as+1)
λ
k−1)zα,k ρ,λ[w(τ
s+1−as+1)ρ](x)
=
∞ X
n=0
σ(n)[w]n kΓk(nρk+λ)
Γk(λ+ρnk)
Γk(λ+ρnk−µ)
(τs+1−as+1)λk+ρn−1
(s+ 1)−µk
= (x
s+1−as+1)λ−kµ−1 (s+ 1)−kµ
zα,kρ,λ−µ[w(x s+1−
as+1)]ρ.
Which complete the desired proof.
3. Properties of the operator skε ω,σ a+;ρ,λf
(x)
Theorem 3.1. For k > 0, ω, σ, λ, ρ ∈ C; R(ω) > 0, R(λ) > 0, R(ρ) > 0, and
R(µ)>0, then the following result holds:
s kε
ω,σ a+;ρ,λ τ
s+1−as+1µk−1
(x) = (x
s+1−as+1)λ+kµ−1Γ k(µ)
(s+ 1) z
σ,k
ρ,λ+µ[w x
s+1−as+1ρ
]. (3.1)
Proof. From (2.1), we have
s kε
ω,σ a+;ρ,λ τ
s+1−as+1µk−1
(x)
= 1 k
Z x
a
τs+1−as+1
µ
k−1(xs+1−τs+1)λk−1
zσ,kρ,λ[ω x s+1−
τs+1ρ]τsf(τ)dτ.
This can be written as
s kε
ω,σ a+;ρ,λ τ
s+1−as+1µk−1
(x)
=
∞ X
n=0
σ(n)[w]n kΓk(nρk+λ)
1 k
Z x
a
xs+1−τs+1
λ+ρkn
k −1 τs+1−as+1 µ
k−1τsdτ
= (x
s+1−αs+1)λ+kµ−1Γ k(µ)
(s+ 1)
∞ X
n=0
σ(n)[ω(xs+1−τs+1)ρ]n
kΓk(nρk+λ)
Γk(nρk+λ)
Γk(nρk+λ+µ)
= (x
s+1−as+1)λ+kµ−1Γ k(µ)
(s+ 1) z
σ,k
ρ,λ+µ[w x
s+1−as+1ρ
].
Which complete the desired proof.
Theorem 3.2. Let s ∈ R\ {−1}; k > 0; ω, λ, ρ ∈ C; R(ω) > 0, R(ω) > 0, R(λ) > 0, R(ρ)>0, R(µ)>0 and x > a, then
(skIaµ+) ksεω,σa+;ρ,λf(x) = 1 (s+ 1)µk
s kε
ω,σ a+;ρ,λ+µf
(x) = skεω,σa+;ρ,λ[skIaµ+f](x),
(3.2)
hold for any f ∈L[ρ, λ].
Proof. From equations (1.11) and (2.1), we have (skIaµ+) skεω,σa+;ρ,λf
(x)
= (s+ 1) 1−µk
kΓk(µ) Z x
a s kε
ω,σ a+;ρ,λf
(τ) (xs+1−τs+1)1−kµ
τsdτ
= (s+ 1) 1−µ
k
k2Γ
k(µ) Z x
a
xs+1−τs+1µk−1
×
Z τ
a
τs+1−us+1
λ k−1
zσ,kρ,λ w τ
s+1−us+1ρ
f(u)usdu
τsdτ.
By changing the order of integration, we obtain (skIaµ+) ksεω,σa+;ρ,λf(x)
= 1 k
Z x
a
(s+ 1)1−µk
kΓk(µ)
Z x
u
xs+1−τs+1
µ
k−1 τs+1−us+1 λ k−1
zσ,kρ,λ w τ
s+1−us+1ρ
τsdτif(u)usdu. (3.3)
Substitutingτs+1−us+1 =ts+1 , this implies τsdτ =tsdt. Therefore (3.3) can be written
as
(skIaµ+) ksεω,σa+;ρ,λf(x)
= 1 k
Z x
a
(s+ 1)1−µk
kΓk(µ) "
Z xs+1−us+1
0
xs+1−ts+1−us+1
µ
k−1 ts+1 λ k−1
zσ,kρ,λ w t s+1ρ
tsdt
i
f(u)usdu.
By the use of (1.11) and applying (2.10), we have (skIaµ+) skεω,σa+;ρ,λf
(x)
= 1
k(s+ 1)µk Z x
a
xs+1−us+1µ+kλ−1
zσ,kρ,λ+µ w x
s+1−us+1ρ
f(u)usdu.
thus, we get
(skIaµ+) ksεω,σa+;ρ,λf(x) = 1 (s+ 1)µk
s kε
ω,σ a+;ρ,λ+µf
To prove the second part, consider the right part of (3.2), we have
s kε
ω,σ a+;ρ,λ[
s kI
µ a+f]
(x) = 1
k
Z x
a
xs+1−τs+1λk−1
zσ,kρ,λ w x
s+1−τs+1ρ
[skIaµ+f]τsdτ
= 1 k
Z x
a
xs+1−τs+1
λ k−1
zσ,kρ,λ
w xs+1−τs+1
ρ k
(s+ 1)1−µ k
kΓk(µ) Z τ
a
f(u)
(τs+1−us+1)1−µk
usdu
!
dτ
By changing the order of integration, we obtain
s kε
ω,σ a+;ρ,λ[
s kI
µ a+f]
(x)
= 1 k
Z x
a
(s+ 1)1−µk
kΓk(µ)
Z x
u
xs+1−τs+1
µ
k−1 τs+1−us+1 λ k−1
zσ,kρ,λ w x
s+1−τs+1ρ
τsdτif(u)usdu. Again making the use of (1.11) and applying (2.10), we have
s kε
ω,σ a+;ρ,λ[
s kI
µ a+f]
(x)
= 1
k(s+ 1)µk Z x
a
xs+1−us+1
µ+λ k −1
zσ,kρ,λ+µ w x
s+1−us+1ρ
f(u)us.
thus, we get
s kε
ω,σ a+;ρ,λ[
s kI
µ a+f]
(x) = 1
(s+ 1)µk s kε
ω,σ a+;ρ,λ+µf
(x).
Which complete the desired proof.
References
[1] A. Atangana, D. Baleanu, New fractional derivatives with nonlocal and non-singular kernel: Theory and application to heat transfer model, Thermal Science 20, 763-769, doi:10.2298/tsci160111018a (2016).
[2] P. Agarwal, M. Jaleli, M. Tomar,Certain Hermite-Hadamard type inequalities via gen-eralized k-fractional integrals, Journal of Inequalities and Applications 2017:55 (2017). [3] R. Diaz, E. Pariguan,On hypergeometric functions and Pochhammer k-symbol, Divulg.
Mat. 17, 179-192 (2007).
[4] A. A. Kilbas, M. Saigo, R. K. Saxena, Generalized Mittag-Lefler function and gen-eralized fractional calculus operators, Integral Transforms and Special Functions 15, 31-49, doi:10.1080/10652460310001600717 (2004).
[6] V. Kiryakova,Generalized fractional calculus and applications, Vol. 38 (Longman Sci-entific and Technical, 1994).
[7] K. S. Miller, B. Ross, An introduction to the fractional calculus and fractional differ-ential equations, (John Wiley and Sons, 1993).
[8] D. S. Mitrinovic, J. E. Pecaric, A. M. Fink,Classical and new inequalities in analysis, Mathematics and its Applications, ( Kluwer Academic Publishers Group, 1993). [9] K. S. Nisar, G. Rahman, D. Baleanu, S. Mubeen, M. Arshad,The (k, s)-fractional
cal-culus of k-Mittag-Leffler function, Advances in Difference Equations (2017) 2017:118. [10] K. S. Nisar, G. Rahman, J. Choi, S. Mubeen, M. Arshad,Generalized hypergeometric k-functions via (k, s)-fractional calculus, J. Nonlinear Sci. Appl., 10 (2017), 17911800. [11] S.Mubeen, G. M. Habibullah,k-Fractional Integrals and Application Int. J. Contemp.
Math. Sciences 7, 89 - 94 (2012).
[12] S. Mubeen, S. Iqbal, Gruss type integral inequalities for generalized Riemann-Liouville k-fractional integrals, Journal of Inequalities and Applications 2016:109 (2016).
[13] S. D. Purohit, S. L. Kalla,On fractional partial differential equations related to quan-tum mechanics, Journal of Physics A: Mathematical and Theoretical 44, 045202 (2011). [14] M. Z. Sarikaya, Z. Dahmani, M. E. Kiris, F. Ahmad, (k; s)-Riemann-Liouville
frac-tional integral and applications, Hacet. J. Math. Stat. 45, 77-89 (2016).
[15] H. M. Srivastava, ˇZ. Tomovski, Fractional calculus with an integral operator con-taining a generalized Mittag-Lefler function in the kernel, Applied Mathematics and Computation 211 (2009), 198-210.
[16] M. Tomar, S. Maden, E. Set, (k, s)-Riemann-Liouville fractional integral inequalities for continuous random variables, Arab. J. Math. 6:55-63 (2017).
[17] T. Tunc, H. Budak, F. Usta, M. Z. Sarikaya, On new generalized frac-tional integral operators and related fracfrac-tional inequalities, ResearchGate, : https://www.researchgate.net/publication/313650587.
G. Rahman: Department of Mathematics, International Islamic University, Islamabad, Pakistan
E-mail address: [email protected]
A. Ghaffar: Department of Mathematical Science, BUITEMS Quetta, Pakistan
E-mail address: [email protected]
K. S. Nisar: Department of Mathematics, College of Arts and Science, Prince Sattam bin Abdulaziz University, Wadi Al dawaser, Riyadh region 11991, Saudi Arabia
E-mail address: [email protected], [email protected]
Azeema: Department of Mathematics, SBK Women’s University, Quetta, Pakistan