The insulated and perfect conductivity problems arising from high-contrast composite materials are considered in all dimensions. The solution and its gradient, respectively, represent the electric potential and field. The novelty of this paper lies in finding exact solutions for the insulated and perfect conductivity problems with concentric balls. Our results show that there appears no electric field concentration for the insulated conductivity problem, while the electric field for the perfect conductivity problem exhibits sharp singularity with respect to the small distance between interfacial boundaries of the interior and exterior balls. This discrepancy reveals that concentric balls is the optimal structure of insulated composites, but not for superconducting composites.
Citation: Zhiwen Zhao. Exact solutions for the insulated and perfect conductivity problems with concentric balls[J]. Mathematics in Engineering, 2023, 5(3): 1-11. doi: 10.3934/mine.2023060
[1] | Moquddsa Zahra, Dina Abuzaid, Ghulam Farid, Kamsing Nonlaopon . On Hadamard inequalities for refined convex functions via strictly monotone functions. AIMS Mathematics, 2022, 7(11): 20043-20057. doi: 10.3934/math.20221096 |
[2] | Ghulam Farid, Hafsa Yasmeen, Hijaz Ahmad, Chahn Yong Jung . Riemann-Liouville Fractional integral operators with respect to increasing functions and strongly (α,m)-convex functions. AIMS Mathematics, 2021, 6(10): 11403-11424. doi: 10.3934/math.2021661 |
[3] | Hengxiao Qi, Muhammad Yussouf, Sajid Mehmood, Yu-Ming Chu, Ghulam Farid . Fractional integral versions of Hermite-Hadamard type inequality for generalized exponentially convexity. AIMS Mathematics, 2020, 5(6): 6030-6042. doi: 10.3934/math.2020386 |
[4] | Maryam Saddiqa, Ghulam Farid, Saleem Ullah, Chahn Yong Jung, Soo Hak Shim . On Bounds of fractional integral operators containing Mittag-Leffler functions for generalized exponentially convex functions. AIMS Mathematics, 2021, 6(6): 6454-6468. doi: 10.3934/math.2021379 |
[5] | Maryam Saddiqa, Saleem Ullah, Ferdous M. O. Tawfiq, Jong-Suk Ro, Ghulam Farid, Saira Zainab . k-Fractional inequalities associated with a generalized convexity. AIMS Mathematics, 2023, 8(12): 28540-28557. doi: 10.3934/math.20231460 |
[6] | Atiq Ur Rehman, Ghulam Farid, Sidra Bibi, Chahn Yong Jung, Shin Min Kang . k-fractional integral inequalities of Hadamard type for exponentially (s,m)-convex functions. AIMS Mathematics, 2021, 6(1): 882-892. doi: 10.3934/math.2021052 |
[7] | Yue Wang, Ghulam Farid, Babar Khan Bangash, Weiwei Wang . Generalized inequalities for integral operators via several kinds of convex functions. AIMS Mathematics, 2020, 5(5): 4624-4643. doi: 10.3934/math.2020297 |
[8] | Manar A. Alqudah, Artion Kashuri, Pshtiwan Othman Mohammed, Muhammad Raees, Thabet Abdeljawad, Matloob Anwar, Y. S. Hamed . On modified convex interval valued functions and related inclusions via the interval valued generalized fractional integrals in extended interval space. AIMS Mathematics, 2021, 6(5): 4638-4663. doi: 10.3934/math.2021273 |
[9] | Moquddsa Zahra, Muhammad Ashraf, Ghulam Farid, Kamsing Nonlaopon . Inequalities for unified integral operators of generalized refined convex functions. AIMS Mathematics, 2022, 7(4): 6218-6233. doi: 10.3934/math.2022346 |
[10] | Wenfeng He, Ghulam Farid, Kahkashan Mahreen, Moquddsa Zahra, Nana Chen . On an integral and consequent fractional integral operators via generalized convexity. AIMS Mathematics, 2020, 5(6): 7632-7648. doi: 10.3934/math.2020488 |
The insulated and perfect conductivity problems arising from high-contrast composite materials are considered in all dimensions. The solution and its gradient, respectively, represent the electric potential and field. The novelty of this paper lies in finding exact solutions for the insulated and perfect conductivity problems with concentric balls. Our results show that there appears no electric field concentration for the insulated conductivity problem, while the electric field for the perfect conductivity problem exhibits sharp singularity with respect to the small distance between interfacial boundaries of the interior and exterior balls. This discrepancy reveals that concentric balls is the optimal structure of insulated composites, but not for superconducting composites.
Fractional calculus is the study of integrals and derivatives of any arbitrary real or complex order. Its origin goes far back in 1695 when Leibniz and l'Hospital started discussion on the meaning of semi-derivative. The 17th century witnessed the attention of this debate by many researchers working in the field of mathematics. Seeking in the formulation of fractional derivative/integral formulas Riemann and Liouville obtained their definitions with complementary functions. Later by Sonin and Letnikov along with others worked out the final form of fractional integral operator named Riemann-Liouville fractional integral/derivative operator, see [1,2]. After this definition of fractional integral/derivative operator the subject of fractional calculus become prominent in generalizing and extending the concepts of calculus and their applications. From the last few years, use of fractional calculus is also followed by scientists from various fields of engineering, sciences and economics. Some fractional integral operators have also been introduced in recent past see [3,4,5]. By applying fractional operators, extensive research has been carried out to establish various inequalities with motivating results, see [6,7,8,9].
Definition 1. [10] Let φ∈L1[ϱ1,ϱ2]. Then left-sided and right-sided Riemann-Liouville fractional integrals of a function φ of order ζ where ℜ(ζ)>0 are given by
Iζϱ+1φ(x)=1Γ(ζ)∫xϱ1(x−t)ζ−1φ(t)dt,x>ϱ1, | (1.1) |
and
Iζϱ−2φ(x)=1Γ(ζ)∫ϱ2x(t−x)ζ−1φ(t)dt,x<ϱ2. | (1.2) |
The k-analogue of the Riemann-Liouville fractional integral is defined as follows:
Definition 2. [11] Let φ∈L1[ϱ1,ϱ2]. Then k-fractional Riemann-Liouville integrals of order ζ where ℜ(ζ)>0, k>0, are given by
kIζϱ+1φ(x)=1kΓk(ζ)∫xϱ1(x−t)ζk−1φ(t)dt,x>ϱ1, | (1.3) |
and
kIζϱ−2φ(x)=1kΓk(ζ)∫ϱ2x(t−x)ζk−1φ(t)dt,x<ϱ2, | (1.4) |
where Γk(.) is defined by [12]:
Γk(ζ)=∫∞0tζ−1e−tkkdt,ℜ(ζ)>0. |
The generalized Riemann-Liouville fractional integrals via a monotonically increasing function are given as follows:
Definition 3. [1] Let φ∈L1[ϱ1,ϱ2]. Also let ψ be an increasing and positive monotone function on (ϱ1,ϱ2], further ψ has a continuous derivative ψ′ on (ϱ1,ϱ2). The left as well as right fractional integral operators of order ζ where ℜ(ζ)>0 of φ with respect to ψ on [ϱ1,ϱ2] are given by
ψIζϱ+1φ(x)=1Γ(ζ)∫xϱ1ψ′(t)(ψ(x)−ψ(t))ζ−1φ(t)dt,x>ϱ1, | (1.5) |
and
ψIζϱ−2φ(x)=1Γ(ζ)∫ϱ2xψ′(t)(ψ(t)−ψ(x))ζ−1φ(t)dt,x<ϱ2. | (1.6) |
The k-analogues of the above generalized Riemann-Liouville fractional integrals are defined as follows:
Definition 4. [13] Let φ∈L1[ϱ1,ϱ2]. Also let ψ be an increasing and positive monotone function on (ϱ1,ϱ2], further ψ has a continuous derivative ψ′ on (ϱ1,ϱ2). Therefore left as well as right k-fractional integral operators of order ζ where ℜ(ζ)>0 of φ with respect to ψ on [ϱ1,ϱ2] are given by
ψkIζϱ+1φ(x)=1kΓk(ζ)∫xϱ1ψ′(t)(ψ(x)−ψ(t))ζk−1φ(t)dt,x>ϱ1, | (1.7) |
and
ψkIζϱ−2φ(x)=1kΓk(ζ)∫ϱ2xψ′(t)(ψ(t)−ψ(x))ζk−1φ(t)dt,x<ϱ2. | (1.8) |
For a detailed study of fractional integrals we refer the readers to [2,14]. Next, we give the definition of exponentially (α,h−m)-convex function as follows:
Definition 5. [15] Let J⊆R be an interval containing (0,1) and let h:J→R be a non-negative function. A function φ:[0,ϱ2]→R is called exponentially (α,h−m)-convex function, if for all x,y∈[0,ϱ2], t∈(0,1), (α,m) ∈(0,1]2 and η∈R one has
φ(tx+m(1−t)y)≤h(tα)φ(x)eηx+mh(1−tα)φ(y)eηy. | (1.9) |
The above definition provides some kinds of exponential convexities as follows:
Remark 1. (i) If we substitute α=1 and h(t)=ts, then exponentially (s,m)-convex function in the second sense introduced by Qiang et al. in [16] can be obtained.
(ii) If we substitute α=m=1 and h(t)=ts, then exponentially s-convex function introduced by Mehreen et al. in [17] can be obtained.
(iii) If we substitute α=m=1 and h(t)=t, then exponentially convex function introduced by Awan et al. in [18] can be obtained.
In [19,20], the following Hadamard inequality for convex function φ:[ϱ1,ϱ2]→R is studied:
φ(ϱ1+ϱ22)≤1ϱ2−ϱ1∫ϱ2ϱ1φ(ν)dν≤φ(ϱ1)+φ(ϱ2)2. | (1.10) |
If this inequality holds in reverse order, then the function f is called concave function. This inequality was first published by Hermite in 1883 and later Hadamard proved it independently in 1893. Since its occurrence, it is in focus of researchers and had/has been studied for different kinds of convex functions. In past two decades it is generalized by using various types of fractional integral operators and authors have investigated a lot of versions of this inequality, see [21,22,23] and references therein. Our aim in this article is to apply the generalized Riemann-Liouville fractional integrals with monotonically increasing function to obtain the Hadamard inequalities for exponentially (α,h−m)-convex functions. Two fractional integral identities are applied to get the error bounds of Hadamard type inequalities. Following two theorems give the fractional versions of Hadamard inequalities for Riemann-Liouville fractional integrals.
Theorem 1. [24] Let φ:[ϱ1,ϱ2]→R be a positive function with 0≤ϱ1<ϱ2 and φ∈L1[ϱ1,ϱ2]. If φ is a convex function on [ϱ1,ϱ2], then the following fractional integral inequality holds:
φ(ϱ1+ϱ22)≤Γ(ζ+1)2(ϱ2−ϱ1)ζ[Iζϱ+1φ(ϱ2)+Iζϱ−2φ(ϱ1)]≤φ(ϱ1)+φ(ϱ2)2, | (1.11) |
with ζ>0.
Theorem 2. [25] Under the assumptions of Theorem 1, the following fractional integral inequality holds:
φ(ϱ1+ϱ22)≤2ζ−1Γ(ζ+1)(ϱ2−ϱ1)ζ[Iζ(ϱ1+ϱ22)+φ(ϱ2)+Iζ(ϱ1+ϱ22)−φ(ϱ1)]≤φ(ϱ1)+φ(ϱ2)2, | (1.12) |
with ζ>0.
The following theorem gives an error estimation of the inequality (1.11).
Theorem 3. [24] Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) with ϱ1<ϱ2. If |φ′| is convex on [ϱ1,ϱ2], then the following fractional integral inequality holds:
|φ(ϱ1)+φ(ϱ2)2−Γ(ζ+1)2(ϱ2−ϱ1)ζ[Iζϱ1+φ(ϱ2)+Iζϱ−2φ(ϱ1)]|≤ϱ2−ϱ12(ζ+1)(1−12ζ)[|φ′(ϱ1)|+|φ′(ϱ2)|]. | (1.13) |
The k-analogues of Theorems 1 and 2 are given in the next two results.
Theorem 4. [26] Let φ:[ϱ1,ϱ2]→R be a positive function with 0≤ϱ1<ϱ2. If φ is a convex function on [ϱ1,ϱ2], then the following inequality for k-fractional integral holds:
φ(ϱ1+ϱ22)≤Γk(ζ+k)2(ϱ2−ϱ1)ζk[kIζϱ+1φ(ϱ2)+kIζϱ−2φ(ϱ1)]≤φ(ϱ1)+f(ϱ2)2, | (1.14) |
with ζ>0.
Theorem 5. [27] Under the assumptions of Theorem 4, the following inequality for k-fractional integral holds:
φ(ϱ1+ϱ22)≤2ζk−1Γk(ζ+k)(ϱ2−ϱ1)ζk[kIζ(ϱ1+ϱ22)+φ(ϱ2)+kIζ(ϱ1+ϱ22)−φ(ϱ1)]≤φ(ϱ1)+φ(ϱ2)2, | (1.15) |
with ζ>0.
An error estimation of the inequality (1.14) is given in the following theorem.
Theorem 6. [26] Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) with 0≤ϱ1<ϱ2. If |φ′| is convex on [ϱ1,ϱ2], then the following inequality for k-fractional integral holds:
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(ϱ2−ϱ1)ζk[kIζϱ1+φ(ϱ2)+kIζϱ−2φ(ϱ1)]|≤ϱ2−ϱ12(ζk+1)(1−12ζk)[|φ′(ϱ1)|+|φ′(ϱ2)|]. | (1.16) |
The rest of the paper is organized as follows: In Section 2, we establish Hadamard inequalities for generalized Riemann-Liouville fractional integrals of exponentially (α,h−m)-convex functions. The special cases, associated with previously published results are presented in terms of their generalizations. In Section 3, using two fractional integral identities, the error bounds of fractional Hadamard inequalities are established. Inequalities proved in this article are in line with the results established in [24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39].
We give two fractional versions of the Hadamard inequality for exponentially (α,h−m)-convex functions, first one is given in the following theorem.
Theorem 7. Let φ:[ϱ1,ϱ2]→R be a positive function with 0≤ϱ1<mϱ2 and φ∈L1[ϱ1,ϱ2]. Also suppose that φ is exponentially (α,h−m)-convex function on [ϱ1,ϱ2], such that ϱ1m, ϱ2m2 ∈[ϱ1,ϱ2]. Then, for η∈R, k>0 and (α,m)∈(0,1]2, the following k-fractional integral inequality holds for operators given in (1.7) and (1.8):
φ(ϱ1+mϱ22)≤Γk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)h(12α)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)H(12)×mζk+1ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζk[g1(η)h(12α)φ(ϱ1)eηϱ1+mg2(η)H(12)φ(ϱ2)eηϱ2]∫10h(tα)tζk−1dt+mζk[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]∫10tζk−1H(t)dt, | (2.1) |
where ζ>0, H(t)=h(1−tα) and
g1(η)={e−ηmϱ2if η<0e−ηϱ1if η>0,g2(η)={e−ηϱ2if η<0e−ηϱ1mif η>0. |
Proof. From exponentially (α,h−m)-convexity of φ, the following inequality holds:
φ(x+my2)≤h(12α)φ(x)eηx+mh(2α−12α)φ(y)eηy. | (2.2) |
By setting x=ϱ1t+m(1−t)ϱ2, y=ϱ1m(1−t)+ϱ2t, t∈[0,1], in (2.2) and integrating the resulting inequality over [0,1] after multiplying with tζk−1, we get
kζφ(ϱ1+mϱ22)≤h(12α)∫10φ(ϱ1t+m(1−t)ϱ2)eη(ϱ1t+m(1−t)ϱ2)tζk−1dt+mH(12)∫10φ(ϱ1m(1−t)+ϱ2t)eη(ϱ1m(1−t)+ϱ2t)tζk−1dt. | (2.3) |
By setting ψ(u)=ϱ1t+m(1−t)ϱ2 and ψ(v)=ϱ1m(1−t)+ϱ2t in (2.3), we obtain
kζφ(ϱ1+mϱ22)≤1(mϱ2−ϱ1)ζk[h(12α)∫ψ−1(mϱ2)ψ−1(ϱ1)φ(ψ(u))eηψ(u)(mϱ2−ψ(u))ζk−1d(ψ(u))+mH(12)∫ψ−1(ϱ2)ψ−1(ϱ1m)φ(ψ(v))eηψ(v)(ψ(v)−ϱ1m)ζk−1d(ψ(v))]. | (2.4) |
Further, taking the maximum value of exponential function and using Definition 4, we get
φ(ϱ1+mϱ22)≤Γk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)h(12α)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1H(12)ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1m))]. |
Again using exponentially (α,h−m)-convexity of φ, for t∈[0,1], we have
g1(η)h(12α)φ(ϱ1t+m(1−t)ϱ2)+mg2(η)H(12)φ(ϱ1m(1−t)+ϱ2t)≤h(tα)[g1(η)h(12α)φ(ϱ1)eηϱ1+mg2(η)H(12)φ(ϱ2)eηϱ2]+mH(t)[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]. | (2.5) |
By integrating (2.5) over the interval [0,1] after multiplying with tζk−1, we get
g1(η)h(12α)∫10φ(tϱ1+m(1−t)ϱ2)tζk−1dt+mg2(η)H(12)∫10φ(ϱ1m(1−t)+tϱ2)tζk−1dt≤[g1(η)h(12α)φ(ϱ1)eηϱ1+mg2(η)H(12)φ(ϱ)eηϱ2]∫10h(tα)tζk−1dt+m[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]∫10tζk−1H(t)dt. | (2.6) |
Again using substitutions as considered in (2.3), the inequality (2.6) leads to the second inequality of (2.1).
The following remark states the connection of Theorem 7 with already established results.
Remark 2. (i) If we substitute η=0, α=1 and ψ(t)=t in (2.1), then [32,Theorem 2.1] can be obtained.
(ii) If we substitute k=m=α=1, η=0 and h(t)=ψ(t)=t in (2.1), then Theorem 1 can be obtained.
(iii) If we substitute η=0, h(t)=ψ(t)=t and m=α=1 in (2.1), then refinement of Theorem 1 can be obtained.
(iv) If we substitute α=k=ζ=m=1, η=0 and h(t)=ψ(t)=t in (2.1), then Hadamard inequality can be obtained.
(v) If we substitute η=0, m=α=1 and h(t)=t in (2.1), then [28,Theorem 1] can be obtained.
(vi) If we substitute η=0, k=m=α=1 and h(t)=t in (2.1), then [37,Theorem 2.1] can be obtained.
(vii) If we substitute α=k=1, h(t)=ψ(t)=t and η=0 in (2.1), then [34,Theorem 2.1] can be obtained.
(viii) If we substitute α=1, h(t)=ts and ψ(t)=t in (2.1), then [39,Theorem 2] can be obtained.
(ix) If we substitute α=1, η=0 and h(t)=ts in (2.1), then [38,Corollary 1] can be obtained.
(x) If we substitute η=0 and k=1 in (2.1), then [31,Corollary 1] can be obtained.
(xi) If we substitute η=0 and α=1 in (2.1), then [31,Corollary 1] gives the refinement of resulting inequality.
In the following we give inequality (2.1) for exponentially (h−m)-convex, exponentially (s,m)-convex, exponentially s-convex, exponentially m-convex and exponentially convex functions.
Corollary 1. If we take α=1 in (2.1), then the following inequality holds for exponentially (h−m)-convex functions:
φ(ϱ1+mϱ22)≤h(12)Γk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1×ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζh(12)k[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ2)eηϱ2]×∫10h(t)tζk−1dt+mζh(12)k[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]∫10tζk−1h(1−t)dt. |
Corollary 2. If we take α=1 and h(t)=ts in (2.1), then the following inequality holds for exponentially (s,m)-convex functions:
φ(ϱ1+mϱ22)≤Γk(ζ+k)2s(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1g2(η)×ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζ2s(ζ+k)[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ1)eηϱ2]+mμB(1+s,μk)2sk[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]. |
If we put m=1 in the above inequality, then the result of exponentially s-convex function can be obtained.
Corollary 3. If we take α=1 and h(t)=t in (2.1), then the following inequality holds for exponentially m-convex functions:
φ(ϱ1+mϱ22)≤Γk(ζ+k)2(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1×ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζ2(ζ+k)[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ2)eηϱ2]+m2(ζ+k)[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]. |
Corollary 4. If we take α=m=1 and h(t)=t in (2.1), then the following inequality holds for exponentially convex functions:
φ(ϱ1+ϱ22)≤Γk(ζ+k)2(ϱ2−ϱ1)ζk×[g1(η)ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(ϱ2))+g2(η)ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]≤ζ2(ζ+k)[g1(η)φ(ϱ1)eηϱ1+g2(η)φ(ϱ2)eηϱ2]+12(ζ+k)[g1(η)φ(ϱ2)eηϱ2+g2(η)φ(ϱ1)eηϱ1]. |
The next theorem is another version of the Hadamard inequality for exponentially (α,h−m)-convex functions.
Theorem 8. Under the assumptions of Theorem 7, the following k-fractional integral inequality holds:
φ(ϱ1+mϱ22)≤2ζkΓk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)h(12α)ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)H(12)mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζk[g1(η)h(12α)φ(ϱ1)eηϱ1+mg2(η)H(12)φ(ϱ2)eηϱ2]∫10h(tα2α)tζk−1dt+ζmk[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]∫10H(t2)tζk−1dt, | (2.7) |
where ζ>0, H(t)=h(1−tα) and
g1(η)={e−ηmϱ2if η<0e−η(ϱ1+mϱ22)if η>0,g2(η)={e−η(ϱ1+mϱ22m)if η<0e−ηϱ1mif η>0. |
Proof. By setting x=ϱ1t2+m(2−t2)ϱ2, y=ϱ1m(2−t2)+ϱ2t2, t∈[0,1], in (2.2) and integrating the resulting inequality over [0,1] after multiplying with tζk−1, we get
kζφ(ϱ1+mϱ22)≤h(12α)∫10φ(ϱ1t2+m(2−t2)ϱ2)eη(ϱ1t2+m(2−t2)ϱ2)tζk−1dt+mH(12)∫10φ(ϱ1m(2−t2)+ϱ2t2)eη(ϱ1m(2−t2)+ϱ2t2)tζk−1dt. | (2.8) |
By setting ψ(u)=ϱ1t2+m(2−t2)ϱ2 and ψ(v)=ϱ1m(2−t2)+ϱ2t2 in (2.8), we obtain
kζφ(ϱ1+mϱ22)≤2ζk(mϱ2−ϱ1)ζk×[h(12α)∫ψ−1(mϱ2)ψ−1(ϱ1+mϱ22)φ(ψ(u))eηψ(u)(mϱ2−ψ(u))ζk−1d(ψ(u))+mζk+1H(12)∫ψ−1(ϱ1+mϱ22m)ψ−1(ϱ1m)φ(ψ(v))eηψ(v)(ψ(v)−ϱ1m)ζk−1d(ψ(v))]. | (2.9) |
Further, taking the maximum value of exponential function and using Definition 4, we get
φ(ϱ1+mϱ22)≤2ζkΓk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)h(12α)ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)h(2α−12α)mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]. |
Again using exponentially (α,h−m)-convexity of φ, for t∈[0,1], we have
g1(η)h(12α)φ(ϱ1t2+m(2−t2)ϱ2)+mg2(η)H(12)φ(ϱ1m(2−t2)+ϱ2t2)≤h(tα2α)[g1(η)h(12α)φ(ϱ1)eηϱ1+mH(12)φ(ϱ2)eηϱ2]+mH(t2)×[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]. | (2.10) |
By integrating (2.10) over [0,1] after multiplying with tζk−1, the following inequality holds
g1(η)h(12α)∫10φ(ϱ1t2+m(2−t2)ϱ2)tζk−1dt+mg2(η)H(12)×∫10φ(ϱ1m(2−t2)+ϱ2t2)tζk−1dt≤[g1(η)h(12α)φ(ϱ1)eηϱ1+mg2(η)H(12)φ(ϱ2)eηϱ2]×∫10h(tα2α)tζk−1dt+m[g1(η)h(12α)φ(ϱ2)eηϱ2+mg2(η)H(12)φ(ϱ1m2)eηϱ1m2]∫10h(2α−tα2α)tζk−1dt. | (2.11) |
Again using substitutions as considered in (2.8), the second inequality of (2.7) can be obtained.
The following remark states the connection of Theorem 8 with already established results.
Remark 3. (i) If we substitute η=0 and k=1 in (2.7), then [31,Corollary 3] can be obtained.
(ii) If we substitute η=0 and α=1 in (2.7), then [31,Theorem 10] gives the refinement of resulting inequality.
(iii) If we take h(t)=ψ(t)=t, m=k=α=1 and η=0 (2.7), then Theorem 2 can be obtained.
(iv) If we substitute h(t)=ψ(t)=t, m=α=1, η=0 in (2.7), then refinement of Theorem 2 can be obtained.
(v) If we use h(t)=ψ(t)=t, m=ζ=k=α=1, η=0 and in (2.7), then the Hadamard inequality can be obtained.
(vi) If we substitute h(t)=t,m=α=1 and η=0 in (2.7), then [35,Corrollary 3] can be obtained.
(vii) If we substitute h(t)=ψ(t=)t,k=α=1 and η=0 in (2.7), then [33,Theorem 2.1] can be obtained.
(viii) If we substitute α=1, h(t)=ts and η=0 in (2.7), then [38,Corollary 3] can be obtained.
(ix) If we substitute h(t)=t, m=α=k=1 and η=0 in (2.7), then [40,Corrollary 3] can be obtained.
In the following we give inequality (2.7) for exponentially (h−m)-convex, exponentially (s,m)-convex, exponentially s-convex, exponentially m-convex and exponentially convex functions.
Corollary 5. If we take α=1 in (2.7), then the following inequality holds for exponentially (h−m)-convex functions:
φ(ϱ1+mϱ22)≤2ζkh(12)Γk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζh(12)k[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ2)eηϱ2]∫10h(t2)tζk−1dt+ζmh(12)k×[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]∫10h(2−t2)tζk−1dt. |
Corollary 6. If we take α=1 and h(t)=ts in (2.7), then the following inequality holds for exponentially (s,m)-convex functions:
φ(ϱ1+mϱ22)≤2ζk−sΓk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζ22s(ζ+sk)[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ2)eηϱ2]+ζmB(1+s,ζk)22sk[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]. |
If we put m=1 in the above inequality, then the result of exponentially s-convex function can be obtained.
Corollary 7. If we take α=1, h(t)=t in (2.7), then the following inequality holds for exponentially m-convex functions:
φ(ϱ1+mϱ22)≤2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+g2(η)mζk+1×ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]≤ζ4(ζ+k)[g1(η)φ(ϱ1)eηϱ1+mg2(η)φ(ϱ2)eηϱ2]+m(ζ+2k)4(ζ+k)[g1(η)φ(ϱ2)eηϱ2+mg2(η)φ(ϱ1m2)eηϱ1m2]. |
Corollary 8. If we take α=m=1, h(t)=t in (2.7), then the following inequality holds for exponentially convex functions:
φ(ϱ1+ϱ22)≤2ζk−1Γk(ζ+k)(ϱ2−ϱ1)ζk[g1(η)ψkIζψ−1(ϱ1+ϱ22)+(φ∘ψ)(ψ−1(ϱ2))+g2(η)ψkIζψ−1(ϱ1+ϱ22)−(φ∘ψ)(ψ−1(ϱ1))]≤ζ4(ζ+k)[g1(η)φ(ϱ1)eηϱ1+g2(η)φ(ϱ2)eηϱ2]+(ζ+2k)4(ζ+k)[g1(η)φ(ϱ2)eηϱ2+g2(η)φ(ϱ1)eηϱ1]. |
In this section we give error estimations of the Hadamard inequalities by using exponentially (α,h−m)-convex functions via generalized Riemann-Liouville fractional integrals. The following identity is useful to prove the next theorem.
Lemma 1. [28] Let ϱ1<ϱ2 and φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2). Also, suppose that φ′∈L1[ϱ1,ϱ2]. Then, for k>0, the following identity holds for the operators given in (1.7) and (1.8):
φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(φ2−φ1)ζk[ψkIζψ−1(φ1)+(φ∘ψ)(ψ−1(φ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]=ϱ2−ϱ12∫10((1−t)ζk−tζk)φ′(tϱ1+(1−t)ϱ2)dt. | (3.1) |
Theorem 9. Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) such that φ′∈L1[ϱ1,ϱ2]. Also suppose that |φ′| is exponentially (α,h−m)-convex on [ϱ1,ϱ2]. Then, for k>0, η∈R and (α,m)∈(0,1]2, the following k-fractional integral inequality holds for the operators given in (1.7) and (1.8):
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12[|φ′(ϱ1)|eηϱ1(∫120h(tα)((1−t)ζk−tζk)dt+∫112h(tα)(tζk−(1−t)ζk)dt)+m|φ′(ϱ2m)|eηϱ2m(∫120H(t)((1−t)ζk−tζk)dt+∫112H(t)(tζk−(1−t)ζk)dt)], | (3.2) |
where H(t) is defined in Theorem 7.
Proof. From Lemma 1, it follows that
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(φ2−φ1)ζk[ψkIζψ−1(φ1)+(φ∘ψ)(ψ−1(φ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12∫10|(1−t)ζk−tζk||φ′(tϱ1+(1−t)ϱ2)|dt. | (3.3) |
By using exponentially (α,h−m)-convexity of |φ′|, for t∈[0,1], we have
|φ′(tϱ1+(1−t)ϱ2)|≤h(tα)|φ′(ϱ1)|eηϱ1+mH(t)|φ′(ϱ2m)|eηϱ2m. | (3.4) |
Using (3.4) in (3.3), we get
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(φ2−φ1)ζk[ψkIζψ−1(φ1)+(φ∘ψ)(ψ−1(φ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12∫10|(1−t)ζk−tζk|[h(tα)|φ′(ϱ1)|eηϱ1+mH(t)|φ′(ϱ2m)|eηϱ2m]=ϱ2−ϱ12[|φ′(ϱ1)|eηϱ1[∫120h(tα)((1−t)ζk−tζk)dt+∫112h(tα)(tζk−(1−t)ζk)dt]+m|φ′(ϱ2m)|eηϱ2m[∫120H(t)((1−t)ζk−tζk)dt+∫112H(t)(tζk−(1−t)ζk)dt]. |
The following remark states the connection of Theorem 9 with already established results.
Remark 4. (i) If we substitute η=0 and k=1 in (3.2), then [31,Corllary 7] can be obtained.
(ii) If we substitute η=0 and α=1 in (3.2), then [31,Theorem 11] gives the refinement of resulting inequality.
(iii) If we substitute h(t)=t, m=α=1 and η=0 in (3.2), then [35,Corollary 10] can be obtained.
(iv) If we substitute m=α=1, η=0 and h(t)=ts in (3.2), then [28,Theorem 2] can be obtained.
(v) If we substitute m=α=1, h(t)=ψ(t)=t and η=0 in (3.2), then [26,Theorem 2.4] can be obtained.
(vi) If we substitute k=m=α=1, h(t)=ψ(t)=t and η=0 in (3.2), then Theorem 3 can be obtained.
(vii) If we substitute k=α=ζ=m=1, h(t)=ψ(t)=t and η=0 in (3.2), then [29,Theorem 2.2] can be obtained.
(viii) If we substitute α=1, η=0 and h(t)=ts in (3.2), then [38,Corollary 5] can be obtained.
In the following we present the inequality (3.2) for exponentially (h−m)-convex, exponentially (s,m)-convex, exponentially s-convex, exponentially m-convex and exponentially convex functions.
Corollary 9. If we take α=1 in (3.2), then the following inequality holds for exponentially (h−m)-convex functions:
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12[|φ′(ϱ1)|eηϱ1(∫120h(t)((1−t)ζk−tζk)dt+∫112h(t)(tζk−(1−t)ζk)dt)+m|φ′(ϱ2m)|eηϱ2m(∫120h(1−t)((1−t)ζk−tζk)dt+∫112h(1−t)(tζk−(1−t)ζk)dt)]. |
Corollary 10. If we take α=1 and h(t)=ts in (3.2), then the following inequality holds for exponentially (s,m)-convex functions:
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12[|φ′(ϱ1)|eηϱ1(2B(12;s+1,ζk+1)+1−(12)s+ζks+ζk+1−B(1+s,1+ζk))+m|φ′(ϱ2m)|eηϱ2m(1−(12)s+ζk+1s+ζk+1−2F1(−s,1+ζk,2+ζk;12)21+ζk(ζk+1)+B(1+s,1+ζk)−k(k+ζ+2s(k+ks+ζ)2F1(−s,1+ζk,2+ζk;12)2s+ζk+1(sk+ζ+k)(ζ+k))]. |
If we put m=1 in the above inequality, then the result of exponentially s-convex function can be obtained.
Corollary 11. If we take α=m=1 and h(t)=t in (3.2), then the following inequality holds for exponentially convex functions:
|φ(ϱ1)+φ(ϱ2)2−Γk(ζ+k)2(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ2)−(φ∘ψ)(ψ−1(ϱ1))]|≤ϱ2−ϱ12(ζk+1)(1−12ζk)(|φ′(ϱ1)|eηϱ1+|φ′(ϱ2)|eηϱ2). |
Corollary 12. If we take α=m=k=1 and h(t)=ψ(t)=t in (3.2), then the following inequality holds for exponentially convex functions via Riemann-Liouville fractional integrals:
|φ(ϱ1)+φ(ϱ2)2−Γ(ζ+1)2(ϱ2−ϱ1)ζ[Iζ(ϱ1)+φ(ϱ2)+Iζ(ϱ2)−φ(ϱ1)]|≤ϱ2−ϱ12(ζk+1)(1−12ζk)(|φ′(ϱ1)|eηϱ1+|φ′(ϱ2)|eηϱ2). |
The following identity will be useful to obtain the next results.
Lemma 2. [38] Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) such that φ∈L1[ϱ1,ϱ2]. Then for k>0 and m∈(0,1], the following integral identity holds for operators given in (1.7) and (1.8):
2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]=mϱ2−ϱ14[∫10tζkφ′(ϱ1t2+m(2−t2)ϱ2)dt−∫10tζkφ′(ϱ1m(2−t2)+ϱ2t2)dt]. | (3.5) |
Theorem 10. Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) such that φ′∈Ł1[ϱ1,ϱ2]. Also suppose that |φ′|q is exponentially (α,h−m)-convex function on [ϱ1,ϱ2] for q≥1. Then, for η∈R, k>0 and (α,m)∈(0,1]2, the following fractional integral inequality holds for operators given in (1.7) and (1.8):
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζk+1)1−1q[(|φ′(ϱ1)|qeηϱ1∫10h(tα2α)tζkdt+m|φ′(ϱ2)|qeηϱ2∫10h(2α−tα2α)tζkdt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10H(t2)tζkdt+|φ′(ϱ2)|qeηϱ2∫10h(tα2α)tζkdt)1q]. | (3.6) |
Proof. We divide the proof in two cases:
Case 1. For q=1. Applying Lemma 2 and using exponentially (α,h−m)-convexity of |φ′|, we have
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14[∫10|tζkφ′(ϱ1t2+m(2−t2)ϱ2)|dt+∫10|tζkφ′(ϱ1m(2−t2)+ϱ2t2)|dt]≤mϱ2−ϱ14[(|φ′(ϱ1)|eηϱ1+|φ′(ϱ2)|eηϱ2)∫10h(tα2α)tζkdt+m(|φ′(ϱ2)|eηϱ2+|φ′(ϱ1m2)|eηϱ1m2)×∫10H(t2)tζkdt]. |
Case 2. Now for q>1. From Lemma 2 and using power mean inequality, we get
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(∫10tζkdt)1−1q[(∫10tζk|φ′(ϱ1t2+m(2−t2)ϱ2)|qdt)1q+(∫10tζk|φ′(ϱ1m(2−t2)+ϱ2t2)|qdt)1q]≤mϱ2−ϱ14(ζk+1)1−1q[(|φ′(ϱ1)|qeηϱ1∫10h(tα2α)tζkdt+m|φ′(ϱ2)|qeηϱ2∫10H(t2)tζkdt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10H(t2)tζkdt+|φ′(ϱ)2|qeηϱ2∫10h(tα2α)tζkdt)1q]. |
Hence (3.6) is obtained.
The following remark states the connection of Theorem 10 with already established results.
Remark 5. (i) If we substitute η=0, α=1 and h(t)=ts in (3.6), then [38,Corollary 7] can be obtained.
(ii) If we substitute h(t)=ψ(t)=t, k=α=1 and η=0 in (3.6), then [33,Theorem 2.4] can be obtained.
(iii) If we substitute η=0, h(t)=ψ(t)=t and α=m=1 in (3.6), then [27,Theorem 3.1] can be obtained.
(iv) If we substitute h(t)=ψ(t)=t, m=k=α=1 and η=0 in (3.6), then [25,Theorem 5] can be obtained.
(v) If we substitute q=m=k=ζ=α=1, h(t)=ψ(t)=t and η=0 in (3.6), then [36,Theorem 2.2] can be obtained.
(vi) If we substitute η=0, m=α=1 and h(t)=t in (3.6), then [35,Corollary 10] can be obtained.
(vii) If we substitute η=0 and k=1 in (3.6), then [31,Corollary 10] can be obtained.
(viii) If we substitute η=0 and α=1 in (3.6), then [31,Theorem 12] gives the refinement of resulting inequality.
In the following we present the inequality (3.6) for exponentially (h−m)-convex, exponentially (s,m)-convex, exponentially s-convex, exponentially m-convex and exponentially convex functions.
Corollary 13. If we take α=1 in (3.6), then the following inequality holds for exponentially (h−m)-convex functions:
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζk+1)1−1q[(|φ′(ϱ1)|qeηϱ1∫10h(t2)tζkdt+m|φ′(ϱ2)|qeηϱ2∫10h(2−t2)tζkdt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10h(2−t2)tζkdt+|φ′(ϱ2)|qeηϱ2∫10h(t2)tζkdt)1q]. |
Corollary 14. If we take α=1 and h(t)=ts in (3.6), then the following inequality holds for exponentially (s,m)-convex functions:
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζk+1)(2s(ζk+s+1))1q[((ζk+1)|φ′(ϱ1)|qeηϱ1+2sm(ζk+s+1)2F1(−s,1+ζk,2+ζk;12)|φ′(ϱ2)|qeηϱ2)1q+(2sm2F1(−s,1+ζk,2+ζk;12)(ζk+s+1)|φ′(ϱ1m2)|qeηϱ1m2+(ζk+1)|φ′(ϱ2)|qeηϱ2)1q]. |
If we put m=1 in the above inequality, then the result of exponentially s-convex function can be obtained.
Corollary 15. If we take α=m=1 and h(t)=t in (3.6), then the following inequality holds for exponentially convex functions:
|2ζk−1Γk(ζ+k)(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+ϱ22)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ1+ϱ22)−(φ∘ψ)(ψ−1(ϱ1))]−φ(ϱ1+ϱ22)|≤ϱ2−ϱ14(ζk+1)(2(ζk+2))1q[((ζk+1)|φ′(ϱ1)|qeηϱ1+(ζk+3)|φ′(ϱ2)|qeηϱ2)1q+((ζk+3)|φ′(ϱ1))|qeηϱ1+(ζk+1)|φ′(ϱ2)|qeηϱ2)1q]. |
Corollary 16. If we take α=k=m=q=ζ=1, h(t)=ψ(t)=t in (3.6), then the following inequality is obtained:
|1(ϱ2−ϱ1)∫10φ(ν)dν−φ(ϱ1+ϱ22)|≤ϱ2−ϱ18[|φ′(ϱ1)|eηϱ1+|φ′(ϱ2)|eηϱ2]. |
Theorem 11. Let φ:[ϱ1,ϱ2]→R be a differentiable mapping on (ϱ1,ϱ2) with 0≤ϱ1<mϱ2. Also suppose that |φ′|q is exponentially (α,h−m)-convex function on [ϱ1,ϱ2] for q>1. Then, for k>0 and (α,m)∈(0,1]2, the following fractional integral inequality holds for the operators given in (1.7) and (1.8):
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζpk+1)1p[(|φ′(ϱ1)|eηϱ1q∫10h(tα2α)dt+m|φ′(ϱ2)|qeηϱ2∫10h(2α−tα2α)dt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10H(t2)dt+|φ′(ϱ2)|qeηϱ2∫10h(tα2α)dt)1q], | (3.7) |
with 1p+1q=1.
Proof. By applying Lemma 2 and using the property of modulus, we get
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14[∫10|tζkφ′(ϱ1t2+m(2−t2)ϱ2)|dt+∫10|tζkφ′(ϱ1m(2−t2)+ϱ2t2)|dt]. |
Now applying Hölder's inequality for integrals, we get
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζpk+1)1p[(∫10|φ′(ϱ1t2+m(2−t2)ϱ2)|qdt)1q+(∫10|φ′(ϱ1m(2−t2)+ϱ2t2)|qdt)1q]. |
Using exponentially (α,h−m)-convexity of |φ′|q, we get
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζpk+1)1p[(|φ′(ϱ1)|eηϱ1q∫10h(tα2α)dt+m|φ′(ϱ2)|qeηϱ2∫10h(2α−tα2α)dt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10h(2α−tα2α)dt+|φ′(ϱ2)|qeηϱ2∫10h(tα2α)dt)1q]. |
The following remark is the connection of Theorem 3.3 with already established results.
Remark 6. (i) If we substitute η=0 and k=1 in (3.7), then [31,Corollary 12] can be obtained.
(ii) If we substitute η=0 and α=1 in (3.7), then [31,Theorem 13] gives the refinement of resulting inequality.
(iii) If we substitute h(t)=ψ(t)=t, k=α=1 and η=0 in (3.7), then [33,Theorem 2.7] can be obtained.
(iv) If we substitute η=0, h(t)=ψ(t)=t and α=m=1 in (3.7), then [27,Theorem 3.2] can be obtained.
(v) If we substitute k=m=ζ=α=1, h(t)=ψ(t)=t and η=0 in (3.7), then [36,Theorem 2.4] can be obtained.
(vi) If we substitute η=0, α=1 and h(t)=ts in (3.7), then [38,Corollary 9] can be obtained.
(vii) If we substitute α=m=1, h(t)=t and η=0 in (3.7), then [35,Corollary 14] can be obtained.
In the following we give inequality (3.7) for exponentially (h−m)-convex, exponentially (s,m)-convex, exponentially s-convex, exponentially m-convex and exponentially convex functions.
Corollary 17. If we take α=1 in (3.7), then the following inequality holds for exponentially (h−m)-convex functions:
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(ζpk+1)1p[(|φ′(ϱ1)|eηϱ1q∫10h(t2)dt+m|φ′(ϱ2)|qeηϱ2∫10h(2−t2)dt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10h(2−t2)dt+|φ′(ϱ2)|qeηϱ2∫10h(t2)dt)1q]. |
Corollary 18. If we take α=1 and h(t)=ts in (3.7), then the following inequality holds for exponentially (s,m)-convex functions:
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−ϱ14(2s(1+s)1q(ζpk+1)1p[(|φ′(ϱ1)|qeηϱ1+2sm|φ′(ϱ2)|q(−1+21+s)eηϱ2)1q+(2sm|φ′(ϱ1m2)|q(−1+21+s)eηϱ1m2+|φ′(ϱ2)|qeηϱ2)1q]. |
If we put m=1 in the above inequality, then the result of exponentially s-convex function can be obtained.
Corollary 19. If we take α=1 and h(t)=t in (3.7), then the following inequality holds for exponentially m-convex functions:
|2ζk−1Γk(ζ+k)(mϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζk+1ψkIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]≤mϱ2−ϱ116(4αpk+1)1p[(|φ′(ϱ1)|qeηϱ1+3m|φ′(ϱ2)|qeηϱ2)1q+(3m|φ′(ϱ1m2)|qeηϱ1m2+|φ′(ϱ2)|qeηϱ2)1q]. |
Corollary 20. If we take α=m=1 and h(t)=t in (3.7), then the following inequality holds for exponentially convex functions:
|2ζk−1Γk(ζ+k)(ϱ2−ϱ1)ζk[ψkIζψ−1(ϱ1+ϱ22)+(φ∘ψ)(ψ−1(ϱ2))+ψkIζψ−1(ϱ1+ϱ22)−(φ∘ψ)(ψ−1(ϱ1))]−φ(ϱ1+ϱ22)|≤ϱ2−ϱ116(4αpk+1)1p[(|φ′(ϱ1)|qeηϱ1+3|φ′(ϱ2)|qeηϱ2)1q+(3|φ′(ϱ1)|qeηϱ1+|φ′(ϱ2)|qeηϱ2)1q]. |
In the following, we give inequality (3.7) for the operators given in (1.5) and (1.6).
Corollary 21. If we take k=1 in (3.7), then the following inequality holds for exponentially convex functions:
|2ζ−1Γ(ζ+1)(mϱ2−ϱ1)ζ[ψIζψ−1(ϱ1+mϱ22)+(φ∘ψ)(ψ−1(mϱ2))+mζ+1ψIζψ−1(ϱ1+mϱ22m)−(φ∘ψ)(ψ−1(ϱ1m))]−12[φ(ϱ1+mϱ22)+mφ(ϱ1+mϱ22m)]|≤mϱ2−a4(ζp+1)1p[(|φ′(ϱ1)|eηϱ1q∫10h(tα2α)dt+m|φ′(ϱ2)|qeηϱ2∫10h(2α−tα2α)dt)1q+(m|φ′(ϱ1m2)|qeηϱ1m2∫10H(t2)dt+|φ′(ϱ2)|qeηϱ2∫10h(tα2α)dt)1q]. |
The Hadamard inequalities presented in this work behave as generalized formulas which generate a number of fractional integral inequalities for all kinds of convex functions connected with exponentially (α,h−m)-convex function. Inequalities for Riemann-Liouville fractional integrals can also be deduced from the results of this paper. This work can be extended for other kinds of fractional integral operators exist in the literature.
We would like to thank reviewers for evaluating the work and giving useful suggestions.
The authors do not have competing interests.
[1] |
H. Ammari, G. Ciraolo, H. Kang, H. Lee, K. Yun, Spectral analysis of the Neumann-Poincaré operator and characterization of the stress concentration in anti-plane elasticity, Arch. Rational Mech. Anal., 208 (2013), 275–304. https://doi.org/10.1007/s00205-012-0590-8 doi: 10.1007/s00205-012-0590-8
![]() |
[2] |
H. Ammari, H. Kang, M. Lim, Gradient estimates for solutions to the conductivity problem, Math. Ann., 332 (2005), 277–286. https://doi.org/10.1007/s00208-004-0626-y doi: 10.1007/s00208-004-0626-y
![]() |
[3] |
H. Ammari, H. Kang, H. Lee, J. Lee, M. Lim, Optimal estimates for the electric field in two dimensions, J. Math. Pure. Appl., 88 (2007), 307–324. https://doi.org/10.1016/j.matpur.2007.07.005 doi: 10.1016/j.matpur.2007.07.005
![]() |
[4] | K. Atkinson, W. M. Han, Spherical harmonics and approximations on the unit sphere: an introduction, Berlin, Heidelberg: Springer, 2012. https://doi.org/10.1007/978-3-642-25983-8 |
[5] |
I. Babuška, B. Andersson, P. Smith, K. Levin, Damage analysis of fiber composites. I. Statistical analysis on fiber scale, Comput. Method. Appl. Mech. Eng., 172 (1999), 27–77. https://doi.org/10.1016/S0045-7825(98)00225-4 doi: 10.1016/S0045-7825(98)00225-4
![]() |
[6] |
E. Bao, Y. Y. Li, B. Yin, Gradient estimates for the perfect conductivity problem, Arch. Rational Mech. Anal., 193 (2009), 195–226. https://doi.org/10.1007/s00205-008-0159-8 doi: 10.1007/s00205-008-0159-8
![]() |
[7] |
E. Bao, Y. Y. Li, B. Yin, Gradient estimates for the perfect and insulated conductivity problems with multiple inclusions, Commun. Part. Diff. Eq., 35 (2010), 1982–2006. https://doi.org/10.1080/03605300903564000 doi: 10.1080/03605300903564000
![]() |
[8] |
E. Bonnetier, F. Triki, On the spectrum of the Poincaré variational problem for two close-to-touching inclusions in 2D, Arch. Rational Mech. Anal., 209 (2013), 541–567. https://doi.org/10.1007/s00205-013-0636-6 doi: 10.1007/s00205-013-0636-6
![]() |
[9] |
G. Ciraolo, A. Sciammetta, Gradient estimates for the perfect conductivity problem in anisotropic media, J. Math. Pure. Appl., 127 (2019), 268–298. https://doi.org/10.1016/j.matpur.2018.09.006 doi: 10.1016/j.matpur.2018.09.006
![]() |
[10] |
G. Ciraolo, A. Sciammetta, Stress concentration for closely located inclusions in nonlinear perfect conductivity problems, J. Differ. Equations, 266 (2019), 6149–6178. https://doi.org/10.1016/j.jde.2018.10.041 doi: 10.1016/j.jde.2018.10.041
![]() |
[11] | H. J. Dong, Y. Y. Li, Z. L. Yang, Optimal gradient estimates of solutions to the insulated conductivity problem in dimension greater than two, arXiv: 2110.11313v2. |
[12] | H. J. Dong, Y. Y. Li, Z. L. Yang, Gradient estimates for the insulated conductivity problem: the non-umbilical case, arXiv: 2203.10081. |
[13] |
Y. Gorb, A. Novikov, Blow-up of solutions to a p-Laplace equation, Multiscale Model. Simul., 10 (2012), 727–743. https://doi.org/10.1137/110857167 doi: 10.1137/110857167
![]() |
[14] | Y. Y. Li, Z. L. Yang, Gradient estimates of solutions to the insulated conductivity problem in dimension greater than two, Math. Ann., in press. https://doi.org/10.1007/s00208-022-02368-x |
[15] |
M. Lim, K. Yun, Blow-up of electric fields between closely spaced spherical perfect conductors, Commun. Part. Diff. Eq., 34 (2009), 1287–1315. https://doi.org/10.1080/03605300903079579 doi: 10.1080/03605300903079579
![]() |
[16] | B. Weinkove, The insulated conductivity problem, effective gradient estimates and the maximum principle, Math. Ann., in press. https://doi.org/10.1007/s00208-021-02314-3 |
[17] |
K. Yun, Estimates for electric fields blown up between closely adjacent conductors with arbitrary shape, SIAM J. Appl. Math., 67 (2007), 714–730. https://doi.org/10.1137/060648817 doi: 10.1137/060648817
![]() |
[18] |
K. Yun, Optimal bound on high stresses occurring between stiff fibers with arbitrary shaped cross-sections, J. Math. Anal. Appl., 350 (2009), 306–312. https://doi.org/10.1016/j.jmaa.2008.09.057 doi: 10.1016/j.jmaa.2008.09.057
![]() |
[19] |
K. Yun, An optimal estimate for electric fields on the shortest line segment between two spherical insulators in three dimensions, J. Differ. Equations, 261 (2016), 148–188. https://doi.org/10.1016/j.jde.2016.03.005 doi: 10.1016/j.jde.2016.03.005
![]() |
1. | Muhammad Bilal Khan, Muhammad Aslam Noor, Nehad Ali Shah, Khadijah M. Abualnaja, Thongchai Botmart, Some New Versions of Hermite–Hadamard Integral Inequalities in Fuzzy Fractional Calculus for Generalized Pre-Invex Functions via Fuzzy-Interval-Valued Settings, 2022, 6, 2504-3110, 83, 10.3390/fractalfract6020083 | |
2. | Soubhagya Kumar Sahoo, Hijaz Ahmad, Muhammad Tariq, Bibhakar Kodamasingh, Hassen Aydi, Manuel De la Sen, Hermite–Hadamard Type Inequalities Involving k-Fractional Operator for (h¯,m)-Convex Functions, 2021, 13, 2073-8994, 1686, 10.3390/sym13091686 | |
3. | Hongling Zhou, Muhammad Shoaib Saleem, Waqas Nazeer, Ahsan Fareed Shah, Hermite-Hadamard type inequalities for interval-valued exponential type pre-invex functions via Riemann-Liouville fractional integrals, 2022, 7, 2473-6988, 2602, 10.3934/math.2022146 | |
4. | Tareq Saeed, Muhammad Bilal Khan, Savin Treanțǎ, Hamed H. Alsulami, Mohammed Sh. Alhodaly, Interval Fejér-Type Inequalities for Left and Right-λ-Preinvex Functions in Interval-Valued Settings, 2022, 11, 2075-1680, 368, 10.3390/axioms11080368 |