ORIGINAL ARTICLE
On the Hadamard’s type inequalities for
co-ordinated convex functions via fractional
integrals
Abdullah Akkurt
a,*, Mehmet Zeki Sar
ıkaya
b, Hu¨seyin Budak
b, Hu¨seyin Y
ıldırım
aaDepartment of Mathematics, Faculty of Science and Arts, University of Kahramanmarasß Su¨tc¸u¨ _Imam, 46100
Kahramanmarasß, Turkey
b
Department of Mathematics, Faculty of Science and Arts, Du¨zce University, Du¨zce, Turkey Received 30 March 2016; accepted 18 June 2016
KEYWORDS Riemann–Liouville fractional integrals; Hadamard’s type Inequalities; Co-ordinated convex functions; Ho¨lder’s inequality
Abstract In this paper, we establish two identities for functions of two variables and apply them to give new Hermite–Hadamard type fractional integral inequalities for double fractional integrals involving functions whose derivatives are bounded or co-ordinates convex function on D :¼ ½a; b ½c; d in R2
with a< b; c < d.
Ó 2016 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Let f: I # R ! R be a convex mapping defined on the interval Iof real numbers and a; b 2 I with a < b. The following double inequality: f aþ b 2 6 1 b a Z b a fðxÞdx 6fðaÞ þ fðbÞ 2 : ð1Þ
is known in the literature as Hermite–Hadamard inequality for convex mappings. Note that some of the classical inequalities for means can be derived from(1)for appropriate particular selections of the mapping f. Both inequalities hold in the reversed direction if f is concave.
It is well known that the Hermite–Hadamard’s inequality plays an important role in nonlinear analysis. Over the last decade, this classical inequality has been improved and gener-alized in a number of ways; there have been a large number of studies on Hermite–Hadamard’s inequality reporting its role in nonlinear analysis (Alomari et al., 2009; Azpeitia, 1994; Bakula and Pecˇaric´, 2004; Dragomir and Pearce, 2000), later, this classical inequality has been improved (Kırmacı and Dikici, 2013; Set et al., 2011; Latif and Dragomir, 2012; Ozdemir et al., 2010) and is generalized in a number of ways (Hussain et al., 2009; Sarikaya and Aktan, 2011; Sarikaya et al., 2014a).
Let us now consider a bidemensional interval
D ¼: a; b½ c; d½ in R2 with a< b and c < d. A mapping
f: D ! R is said to be convex on D if the following inequality: fðtx þ 1 tð Þz; ty þ 1 tð ÞwÞ 6 tf x; yð Þ þ 1 tð Þf z; wð Þ ð2Þ holds, for all ðx; yÞ; z; wð Þ 2 D and t 2 0; 1½ . A function f: D ! R is said to be convex on the co-ordinates on D if the partial mappings fy: a; b½ ! R, fyð Þ ¼ f u; yu ð Þ and * Corresponding author.
E-mail address:abdullahmat@gmail.com(A. Akkurt). Peer review under responsibility of King Saud University.
Production and hosting by Elsevier
Journal of King Saud University – Science (2016) xxx, xxx–xxx
King Saud University
Journal of King Saud University –
Science
www.ksu.edu.sa
www.sciencedirect.com
fx: c; d½ ! R, fxð Þ ¼ f x; vv ð Þ are convex where defined for all x2 a; b½ and y 2 c; d½ (Dragomir and Pearce, 2000).
A formal definition for co-ordinated convex function may be stated as follows:
Definition 1. A function f: D ! R will be called co-ordinated convex on D, for all t; s 2 ½0; 1 and ðx; yÞ; ðu; wÞ 2 D, if the following inequality holds:
fðtx þ 1 tð Þy; su þ 1 sð ÞwÞ 6 tsfðx; uÞ þ sð1 tÞfðy; uÞ
þ tð1 sÞfðx; wÞ þ ð1 tÞð1 sÞfðy; wÞ: ð3Þ
Clearly, every convex function is co-ordinated convex. Furthermore, there exist co-ordinated convex function which is not convex (Dragomir, 2001). Several recent studies have expressed concerns on Hermite–Hadamard’s inequality for some convex function on the co-ordinates on a rectangle from the planeR2(Sarikaya and Yaldiz, 2013; Ozdemir et al., 2011; Sarikaya et al., 2012; Sarikaya et al. (2014c)). More details, one can consult Sarikaya (2014), Sarikaya et al. (2014b) and Sarikaya (2015).
Earlier,Dragomir (2001)establish the following inequality of Hermite–Hadamard type for co-ordinated convex mapping on a rectangle from the planeR2. Later, another proof of a special version of the following theorem, using the definition of the co-ordinated convex function was reported (Sarikaya and Yaldiz, 2013).
Theorem 2. Suppose that f: D ! R is co-ordinated convex on D. Then one has the inequalities:
f aþ b 2 ; cþ d 2 61 2 1 b a Zb a f x;cþ d 2 dxþ 1 d c Z d c f aþ b 2 ; y dy 6 1 b a ð Þ d cð Þ Zb a Zd c fðx; yÞdydx 61 4 1 b a Zb a fðx; cÞdx þ 1 b a Zb a fðx; dÞdx þ 1 d c Zd c fða; yÞdy þ 1 d c Zd c fðb; yÞdy 6fða; cÞ þ fða; dÞ þ fðb; cÞ þ fðb; dÞ 4 ð4Þ
The above inequalities are sharp.
In the following section, some relevant definitions and mathematical preliminaries of fractional calculus theory are presented. For more details, one can consult Gorenflo and Mainardi (1997), Kilbas et al. (2006), Samko et al. (1993), Miller and Ross (1993).
Definition 3. Let f2 L1½a; b. The Riemann–Liouville integrals
Jaaþfand Jabfof ordera > 0 with a P 0 are defined by
JaaþfðxÞ ¼ 1 C að Þ Z x a x t ð Þa1fðtÞdt; x > a; ð5Þ JabfðxÞ ¼ 1 C að Þ Z b x t x ð Þa1fðtÞdt; x < b ð6Þ
respectively. Here,C að Þ is the Gamma function.
It is remarkable thatSarikaya et al. (2012)first give the fol-lowing interesting integral inequalities of Hermite–Hadamard type involving Riemann–Liouville fractional integrals. Theorem 4. Let f: a; b½ ! R be a positive function with 06 a < b and f 2 L1½a; b. If f is a convex function on ½a; b,
then the following inequalities for fractional integrals hold:
f aþ b 2 6C a þ 1ð Þ 2 bð aÞa J a aþfðbÞ þ J a bfðaÞ 6fðaÞ þ fðbÞ 2 ð7Þ witha > 0.
Meanwhile, Sarikaya et al. (2012)presented the following important integral identity including the first-order derivative of f to establish many interesting Hermite–Hadamard type inequalities for convexity functions via Riemann–Liouville fractional integrals of the ordera > 0.
Lemma 5. Let f: a; b½ ! R be a differentiable mapping on a; b
ð Þ with a < b. If f02 L
1½a; b, then the following equality for
fractional integrals holds: fðaÞ þ fðbÞ 2 C a þ 1ð Þ 2 bð aÞa J a aþfðbÞ þ JabfðaÞ ð8Þ ¼b a 2 Z 1 0 1 t ð Þa ta ½ f0ðta þ ð1 tÞbÞdt: ð9Þ
Definition 6. Let f2 L1ð½a; b c; d½ Þ. The Riemann–Liouville
integrals Ja;baþ;cþ; Ja;baþ;d; Ja;bb;cþ and Ja;bb;d of ordera; b > 0 with
a; c P 0 are defined by Ja;baþ;cþfðx; yÞ ¼ 1 C að ÞC bð Þ Z x a Z y c x t ð Þa1ðy sÞb1fðt; sÞdsdt; ð10Þ Ja;baþ;dfðx; yÞ ¼ 1 C að ÞC bð Þ Z x a Z d y x t ð Þa1ðs yÞb1fðt; sÞdsdt; ð11Þ Ja;bb;cþfðx; yÞ ¼ 1 C að ÞC bð Þ Z b x Z y c t x ð Þa1ðy sÞb1fðt; sÞdsdt; ð12Þ and Ja;bb;dfðx; yÞ ¼ 1 C að ÞC bð Þ Z b x Z d y t x ð Þa1ðs yÞb1fðt; sÞdsdt; ð13Þ respectively. Similar to Definitions 3and6we introduce the following fractional integrals:
Jaaþf x; cþ d 2 ¼ 1 C að Þ Z x a x t ð Þa1f t;cþ d 2 dt; ð14Þ Jabf x; cþ d 2 ¼C að Þ1 Z b x t x ð Þa1f t;cþ d 2 dt; ð15Þ Jbcþf aþ b 2 ; y ¼C bð Þ1 Z y c y s ð Þb1f aþ b 2 ; s ds; ð16Þ
Jbdf aþ b 2 ; y ¼C bð Þ1 Z d y s y ð Þb1f aþ b 2 ; s ds: ð17Þ
Objective of the present study is to state and prove the Hermite–Hadamard type inequality for co-ordinated convex mapping on a rectangle from the planeR2. In order to achieve our goal, we first give two important identities and then by using these identities we prove some integral inequalities. We have obtained some results which are a simpler proof of the results presented bySarikaya (2012).
2. Main results
To establish our main results, we need the following first identity:
Lemma 7. Let f: D # R2! R be a partial differentiable mapping on D :¼ ½a; b ½c; d in R2 with a< b and c < d and
frs2 LðDÞ. Then the following equality holds: 4C a þ 1ð ÞC b þ 1ð Þ
b a
ð Þa
d c
ð Þb Ja;baþ;cþfðb;dÞ þ Jaa;bþ;dfðb;cÞ þ Ja;bb;cþfða;dÞ
h þJa;b b;dfða;cÞ i 2C a þ 1ð Þ b a ð Þa Jaaþfðb;cÞ þ J a aþfðb;dÞ þJa bfða;cÞ þ J a bfða;dÞ 2C b þ 1ð Þ d c ð Þb Jbcþfða;dÞ þ J b cþfðb;dÞ þJb dfða;cÞ þ J b dfðb;cÞ þ F ¼ ab b a ð Þaðd cÞb Z b a Z d c b x ð Þa1ðd yÞb1 h þ b xð Þa1 y c ð Þb1þ x að Þa1ðy cÞb1 þ x að Þa1ðd yÞb1i Iðx;yÞdydxo ð18Þ where Iðx; yÞ ¼ Z x a Z y c frsðr; sÞdsdr þ Z x a Z y d frsðr; sÞdsdr þ Z x b Z y c frsðr; sÞdsdr þ Z x b Z y d frsðr; sÞdsdr; ð19Þ and F¼ fða; cÞ þ fða; dÞ þ fðb; cÞ þ fðb; dÞ: ð20Þ
Proof. For any x; t 2 a; b½ and s; y 2 c; d½ ; x – t; s – y, we have Z x t Z y s frsðr; sÞdsdr ¼ Z x t frðr; yÞ frðr; sÞ ½ dr ¼ f r; y½ð Þ f r; sð Þjx t ¼ fðx; yÞ fðx; sÞ fðt; yÞ þ fðt; sÞ: ð21Þ Choose t¼ a; s ¼ c; t ¼ a; s ¼ d; t ¼ b; s ¼ c; t ¼ b; s ¼ d in(21), respectively, we get I1¼ Z x a Z y c
frsðr; sÞdsdr ¼ fðx; yÞ fðx; cÞ fða; yÞ þ fða; cÞ; ð22Þ I2¼ Z x a Z y d
frsðr; sÞdsdr ¼ fðx; yÞ fðx; dÞ fða; yÞ þ fða; dÞ; ð23Þ I3¼ Z x b Z y c frsðr; sÞdsdr ¼ fðx; yÞ fðx; cÞ fðb; yÞ þ fðb; cÞ; ð24Þ and I4¼ Z x b Z y d frsðr; sÞdsdr ¼ fðx; yÞ fðx; dÞ fðb; yÞ þ fðb; dÞ: ð25Þ Adding these four integrals side by side, we obtain
Iðx; yÞ ¼ I1þ I2þ I3þ I4
¼ 4fðx; yÞ 2 fðx; cÞ þ fðx; dÞ½ 2 fða; yÞ þ fðb; yÞ½ þ fða; cÞ þ fða; dÞ þ fðb; cÞ þ fðb; dÞ: ð26Þ Multiplying(26)byðbx4C aÞa1ð ÞC bðdyð ÞÞb1and integrating the resulting equality with respect toðx; yÞ on ½a; b ½c; d, we have
1 4C að ÞC bð Þ Z b a Z d c bx ð Þa1ðdyÞb1Iðx;yÞdydx ¼C að ÞC b1ð Þ Z b a Z d c bx ð Þa1ðdyÞb1fðx;yÞdydx 1 2C að ÞC bð Þ Z b a Z d c bx ð Þa1ðdyÞb1½fðx;cÞþfðx;dÞdx 1 2C að ÞC bð Þ Z b a Z d c bx
ð Þa1ðdyÞb1½fða;yÞþfðb;yÞdydx
þ F 4C að ÞC bð Þ Z b a Z d c bx
ð Þa1ðdyÞb1dydx:
ð27Þ Thus, in(27)by means of simple calculations, we have
Ja;baþ;cþðb; dÞ d c ð Þb 2C b þ 1ð Þ J a aþfðb; cÞ þ Jaaþfðb; dÞ ðb aÞa 2C a þ 1ð Þ J b cþfða; dÞ þ Jbcþfðb; dÞ þC a þ 1ððb aÞaÞC b þ 1ðdð cÞbÞF ¼ 1 4C að ÞC bð Þ Z b a Z d c b x ð Þa1ðd yÞb1Iðx; yÞdydx: ð28Þ
Multiplying(26) byðbx4C aÞa1ð ÞC bðycð ÞÞb1 and integrating the resulting equality with respect toðx; yÞ on ½a; b ½c; d, and by similar calculations, we have Ja;baþ;dfðb; cÞ d c ð Þb 2C b þ 1ð Þ Jaaþfðb; cÞ þ J a aþfðb; dÞ 2C a þ 1ðbð aÞaÞ Jbdfða; cÞ þ Jbdfðb; cÞ þ4C a þ 1ðbð aÞaÞC b þ 1ðdð cÞbÞF ¼4C að ÞC b1 ð Þ Z b a Z d c b x ð Þa1ðy cÞb1Iðx; yÞdydx: ð29Þ
Multiplying(26) byðxa4C aÞa1ð ÞC bðycð ÞÞb1 and integrating the resulting equality with respect toðx; yÞ on ½a; b ½c; d, we have
Ja;bb;cþfða; dÞ d c ð Þb 2C b þ 1ð Þ J a bfða; cÞ þ Jabfða; dÞ ðb aÞ a 2C a þ 1ð Þ J b cþfða; dÞ þ Jbcþfðb; dÞ þ ðb aÞ a d c ð Þb 4C a þ 1ð ÞC b þ 1ð ÞF ¼ 1 4C að ÞC bð Þ Z b a Z d c x a ð Þa1ðy cÞb1Iðx; yÞdydx: ð30Þ
Multiplying(26)byðxa4C aÞa1ð ÞC bðdyð ÞÞb1and integrating the resulting equality with respect toðx; yÞ on ½a; b ½c; d, we have Ja;bb;dfða; cÞ ðd cÞ b 2C b þ 1ð Þ J a bfða; cÞ þ Jabfða; dÞ ðb aÞ a 2C a þ 1ð Þ J b dfða; cÞ þ J b dfðb; cÞ þ ðb aÞ a d c ð Þb 4C a þ 1ð ÞC b þ 1ð ÞF ¼ 1 4C að ÞC bð Þ Z b a Z d c x a ð Þa1ðd yÞb1Iðx; yÞdydx: ð31Þ
Adding these(28)–(31)side by side, which completes the proof.
Corollary 8. If we takea ¼ b ¼ 1 inLemma7, we get 4 b a ð Þ d cð Þ Z b a Z d c fðx;yÞdydx 2 b a ð Þ Z b a f xð ;cÞ þ f x;dð Þ ½ dx ðd c2 Þ Z d c f að ;yÞ þ f b;yð Þ ½ dy þ F ¼ðb a1 Þ d cð Þ Z b a Z d c Iðx;yÞdydx: ð32Þ
Theorem 9. Let f: D # R2! R be a partial differentiable mapping onD :¼ ½a; b ½c; d in R2 with a< b and c < d and frs2 LðDÞ. If frs2 L1ðDÞ, i.e fjj jrsj1¼ sup r;s ð Þ2 a;bð Þ c;dð Þ @2fðr;sÞ @r@s < 1, then one has the inequality:
4C a þ 1ð ÞC b þ 1ð Þ b a ð Þa d c ð Þb Ja;baþ;cþfðb; dÞ þ J a;b aþ;dfðb; cÞ þ J a;b b;cþfða; dÞ h þJa;b b;dfða; cÞ i 2C a þ 1ð Þ b a ð Þa Jaaþfðb; cÞ þ J a aþfðb; dÞ þJa bfða; cÞ þ J a bfða; dÞ 2C b þ 1ð Þ d c ð Þb Jbcþfða; dÞ þ J b cþfðb; dÞ þJb dfða; cÞ þ J b dfðb; cÞ þ F 6 4 fjj jrsj1ðb aÞ d cð Þ: ð33Þ
Proof. FromLemma 7, taking the modulus, it follows that
J
j j ¼ 4C a þ 1ð ÞC b þ 1ð Þ b a
ð Þa
d c
ð Þb Ja;baþ;cþfðb;dÞ þ Jaa;bþ;dfðb;cÞ þ Ja;bb;cþfða;dÞ
h þJa;b b;dfða;cÞ i 2C a þ 1ð Þ b a ð Þa Jaaþfðb;cÞ þ J a aþfðb;dÞ þ J a bfða;cÞ þJa bfða;dÞ 2C b þ 1ð Þ d c ð Þb Jbcþfða;dÞ þ J b cþfðb;dÞ þ J b dfða;cÞ þJb dfðb;cÞ þ F ð34Þ 6 ab b a ð Þa d c ð Þb Z b a Z d c b x ð Þa1ðd yÞb1 h þ b xð Þa1ðy cÞb1þ x að Þa1ðy cÞb1 þ x að Þa1 d y ð Þb1i Z x a Z y c frsðr; sÞ j jdsdr þ Z x a Z d y frsðr; sÞ j jdsdrþ Z b x Z y c frsðr; sÞ j jdsdr þ Z b x Z d y frsðr; sÞ j jdsdr dydx : ð35Þ Since frs2 L1ðDÞ, we get J j j 6 ab fjj jrsj1 b a ð Þa d c ð Þb Zb a Zd c b x ð Þa1ðd yÞb1 Zb a Z d c dsdr dydx þ Zb a Zd c b x ð Þa1ðy cÞb1 Zb a Zd c dsdr dydx ð36Þ þ Z b a Z d c x a ð Þa1ðy cÞb1 Z b a Z d c dsdr dydx þ Z b a Z d c x a ð Þa1ðd yÞb1 Z b a Z d c dsdr dydx ¼ ab fjj jrsj1 b a ð Þa d c ð Þb 4 bð aÞaþ1 a d c ð Þbþ1 b ¼ 4 fjj jrsj1ðb aÞ d cð Þ: ð37Þ
This completes the proof.
Corollary 10. If we takea ¼ b ¼ 1 inTheorem9, we get 4 ba ð Þ dcð Þ Z b a Z d c fðx;yÞdydx 2 ba ð Þ Z b a f xð ;cÞþf x;dð Þ ½ dx 2 dc ð Þ Z d c f að ;yÞþf b;yð Þ ½ dyþF 6 4 fjj jrsj1ðbaÞ dcð Þ: ð38Þ Theorem 11. Let f: D # R2! R be a partial differentiable mapping on D :¼ ½a; b ½c; d in R2 with a< b and c < d and frs2 LðDÞ. If fj j is a convex function on the co-ordinates onrs D, then the following inequality holds:
4C a þ 1ð ÞC b þ 1ð Þ b a
ð Þaðd cÞb Ja;baþ;cþfðb; dÞ þ Ja;baþ;dfðb; cÞ þ Ja;bb;cþfða; dÞ
h þJa;b b;dfða; cÞ i 2C a þ 1ð Þ b a ð Þa Jaaþfðb; cÞ þ Jaaþfðb; dÞ þJa bfða; cÞ þ Jabfða; dÞ 2C b þ 1ð Þ d c ð Þb J b cþfða; dÞ þ J b cþfðb; dÞ þJb dfða; cÞ þ J b dfðb; cÞ þ F 6 b a ð Þ d cð Þ f½jrsða; cÞþj fjrsða; dÞj þ fjrsðb; cÞj þ fjrsðb; dÞj ð39Þ
Proof. Since fjrsðr; sÞj is co-ordinates on D, we know that x2 a; b½ ; y 2 c; d½
frsðr;sÞ j j ¼ frs b r b aaþ r a b ab; d s d ccþ s c d cd 6b r b a d s d cjfrsða;cÞj þ b r b a s c d cjfrsða;dÞj þr a b a d s d cjfrsðb;cÞj þ r a b a s c d cjfrsðb;dÞj: ð40Þ FromLemma 7, we have
J j j 6 ab b a ð Þaðd cÞb Z b a Z d c b x ð Þa1 d y ð Þb1 h þ b xð Þa1 y c ð Þb1þ x að Þa1 y c ð Þb1 þ x að Þa1ðd yÞb1i Z x a Z y c frsðr; sÞ j jdsdr þ Z x a Z d y frsðr; sÞ j jdsdrþ Z b x Z y c frsðr; sÞ j jdsdr þ Z b x Z d y frsðr; sÞ j jdsdr dydx ð41Þ By using co-ordinated convexity of fj j, we getrs
J j j 6 ab b a ð Þaðd cÞb Z b a Z d c b x ð Þa1 d y ð Þb1 h þ b xð Þa1 y c ð Þb1þ x að Þa1 y c ð Þb1 þ x að Þa1ðd yÞb1i Z x a Z y c b r b a d s d cjfrsða;cÞj þb r b a s c d cjfrsða;dÞj þ r a b a d s d cjfrsðb;cÞj þr a b a s c d cjfrsðb;dÞj i dsdr þ Z x a Z d y b r b a d s d cjfrsða;cÞj þb r b a s c d cjfrsða;dÞj þ r a b a d s d cjfrsðb;cÞj þr a b a s c d cjfrsðb;dÞj i dsdr ð42Þ þ Z b x Z y c b r b a d s d cjfrsða;cÞj þ b r b a s c d cjfrsða;dÞj þr a b a d s d cjfrsðb;cÞj þ r a b a s c d cjfrsðb;dÞj dsdr þ Z b x Z d y b r b a d s d cjfrsða;cÞj þ b r b a s c d cjfrsða;dÞj þr a b a d s d cjfrsðb;cÞjþ r a b a s c d cjfrsðb;dÞj dsdr dydx ð43Þ ¼ ab b a ð Þaðd cÞb Z b a Z d c b x ð Þa1ðd yÞb1 h þ b xð Þa1ðy cÞb1þ x að Þa1ðy cÞb1 þ x að Þa1 d y ð Þb1i Z b a Z d c b r b a d s d cjfrsða;cÞj þb r b a s c d cjfrsða;dÞj þr ab a d s d cjfrsðb;cÞj þr a b a s c d cjfrsðb;dÞj i dsdr o dydx¼ A1þ A2þ A3þ A4: ð44Þ
With a simple calculation, we have
A1¼ ab ba ð Þaþ1ðdcÞbþ1 Zb a Zd c bx ð Þa1ðd yÞb1 (Z b a Zd c b r ð Þ d sð Þ fjrsða;cÞj þ b rð Þ s cð Þ fjrsða;dÞj ½ þ r að Þ d sð Þ fjrsðb;cÞjþ r að Þ s cð Þ fjrsðb;dÞjdsdr ) dydx ð45Þ ¼ ab b a ð Þaþ1ðd cÞbþ1 ðb aÞaþ2 2a d c ð Þbþ2 2b ½jfrsða; cÞþj fjrsða; dÞj ( þ fjrsðb; cÞj þ fjrsðb; dÞj ¼ b a ð Þ d cð Þ 4 ½jfrsða; cÞþj fjrsða; dÞj þ fjrsðb; cÞj þ fjrsðb; dÞj: ð46Þ
Similarly, we also have the following equalities
A2¼ ab b a ð Þaþ1ðd cÞbþ1 Z b a Z d c b x ð Þa1ðy cÞb1 Z b a Z d c b r ð Þ d sð Þ fjrsða;cÞj þ b rð Þ s cð Þ fjrsða;dÞj ½ þ r að Þ d sð Þ fjrsðb;cÞj þ r að Þ s cð Þ fjrsðb;dÞjdsdr dydx ¼ðb aÞ d cð Þ 4 ½jfrsða;cÞþj fjrsða;dÞj þ fjrsðb;cÞj þ fjrsðb;dÞj; ð47Þ A3¼ ab b a ð Þaþ1ðd cÞbþ1 Z b a Z d c x a ð Þa1ðy cÞb1 Z b a Z d c b r ð Þ d sð Þ fjrsða;cÞj þ b rð Þ s cð Þ fjrsða;dÞj ½ þ r að Þ d sð Þ fjrsðb;cÞj þ r að Þ s cð Þ fjrsðb;dÞjdsdr dydx ¼ðb aÞ d c4ð Þ½jfrsða;cÞþj fjrsða;dÞj þ fjrsðb;cÞj þ fjrsðb;dÞj ð48Þ and A4¼ ab b a ð Þaþ1ðd cÞbþ1 Z b a Z d c x a ð Þa1ðd yÞb1 Z b a Z d c b r ð Þ d sð Þ fjrsða;cÞj þ b rð Þ s cð Þ fjrsða;dÞj ½ þ r að Þ d sð Þ fjrsðb;cÞj þ r að Þ s cð Þ fjrsðb;dÞjdsdr dydx ¼ðb aÞ d c4ð Þ½jfrsða;cÞþj fjrsða;dÞj þ fjrsðb;cÞj þ fjrsðb;dÞj: ð49Þ
Adding these(46)–(49)side by side, if we put in(44), we obtain
(39). This completes the proof of the theorem.
Corollary 12. If we takea ¼ b ¼ 1 inTheorem11, we get 4 ba ð Þ dcð Þ Z b a Z d c fðx;yÞdydx 2 ba ð Þ Z b a f xð ;cÞþf x;dð Þ ½ dx ðdc2 Þ Z d c f að ;yÞþf b;yð Þ ½ dyþF 6 bað Þ dcð Þ fj½rsð Þa;c þj fjrsða;dÞjþ fjrsð Þb;cjþ fjrsðb;dÞj: ð50Þ
Lemma 13. Let f: D # R2! R be a partial differentiable map-ping on D :¼ ½a; b ½c; d in R2 with a< b; c < d and frs2 LðDÞ. Then the following equality holds:
f aþ b 2 ; cþ d 2 C b þ 1ð Þ 2 dð cÞb J b cþf aþ b 2 ;d þ Jb df aþ b 2 ;c C a þ 1ð Þ 2 bð aÞa J a bf a; cþ d 2 þ Ja aþf b; cþ d 2 þC a þ 1ð ÞC b þ 1ð Þ 4 b að Þaðd cÞb Ja;baþ;cþfðb;dÞ þ J a;b aþ;dfðb;cÞ þ J a;b b;cþfða;dÞ h þJa;b b;dfða;cÞ i ¼ ab 4 bð aÞaðd cÞb Z b a Z d c b t ð Þa1 h n þ t að Þa1i ðd sÞb1þ s cð Þb1 h i Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr !) dsdt: ð51Þ
Proof. Choose x¼aþb2 and y¼cþd2 in(21), we have Z aþb 2 t Z cþd 2 s frsðr; sÞdsdr ¼ f aþ b 2 ; cþ d 2 f aþ b 2 ; s f t;cþ d 2 þ fðt; sÞ: ð52Þ
Multiplying(52)byðbtC aÞa1ð ÞC bðdsð ÞÞb1 and integrating the resulting equality with respect toðs; tÞ on ½a; b ½c; d, we get
1 C að ÞC bð Þ Z b a Z d c b t ð Þa1ðd sÞb1 Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr ( ) dsdt ¼f aþb2 ;cþd2 C að ÞC bð Þ Z b a Z d c b t ð Þa1ðd sÞb1dsdt C að ÞC b1ð Þ Z b a Z d c b t ð Þa1ðd sÞb1f aþ b 2 ;s dsdt C að ÞC b1ð Þ Z b a Z d c b t ð Þa1ðd sÞb1f t;cþ d 2 dsdt þC að ÞC b1ð Þ Z b a Z d c b t ð Þa1ðd sÞb1fðt;sÞdsdt: ð53Þ By simple calculations, we have
b a ð Þaðd cÞb C a þ 1ð ÞC b þ 1ð Þf aþ b 2 ; cþ d 2 ðb aÞa C a þ 1ð ÞJ b cþf aþ b 2 ;d ðd cÞb C b þ 1ð ÞJaaþf b; cþ d 2 þ Ja;b aþ;cþfðb;dÞ ¼ 1 C að ÞC bð Þ Z b a Z d c b t ð Þa1 d s ð Þb1 Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr ( ) dsdt: ð54Þ
Multiplying (52) by ðbtC aÞa1ð ÞC bðscð ÞÞb1, integrating the resulting equality with respect toðs; tÞ on ½a; b ½c; d, and by similar methods above we have
b a ð Þaðd cÞb C a þ 1ð ÞC b þ 1ð Þf aþ b 2 ; cþ d 2 C a þ 1ðbð aÞaÞJbdf aþ b 2 ;c ðd cÞb C b þ 1ð ÞJaaþf b; cþ d 2 þ Ja;b aþ;dfðb;cÞ ¼ 1 C að ÞC bð Þ Z b a Z d c b t ð Þa1 d s ð Þb1 Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr ( ) dsdt: ð55Þ
Multiplying (52) by ðtaC aÞa1ð ÞC bðdsð ÞÞb1 integrating the resulting equality with respect toðs; tÞ on ½a; b ½c; d, and by similar methods above we have
b a ð Þaðd cÞb C a þ 1ð ÞC b þ 1ð Þf aþ b 2 ; cþ d 2 C a þ 1ðb að ÞaÞJbcþf aþ b 2 ;d C b þ 1ðd cð ÞbÞJabf a; cþ d 2 þ Ja;b b;cþfða;dÞ ¼C að ÞC b1ð ÞZ b a Z d c b t ð Þa1ðd sÞb1 Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr ( ) dsdt: ð56Þ Multiplying (52) by ðtaC aÞa1ð ÞC bðscð ÞÞb1 integrating the resulting equality with respect toðs; tÞ on ½a; b ½c; d, and by similar methods above we have
b a ð Þaðd cÞb C a þ 1ð ÞC b þ 1ð Þf aþ b 2 ; cþ d 2 C a þ 1ðb að ÞaÞJbdf aþ b 2 ;c C b þ 1ðd cð ÞbÞJabf a; cþ d 2 þ Ja;b b;dfða;cÞ ¼C að ÞC b1ð Þ Z b a Z d c b t ð Þa1ðd sÞb1 Z aþb 2 t Z cþd 2 s frsðr;sÞdsdr ( ) dsdt: ð57Þ
Adding these(54)–(57)side by side and multiplying both sides by4 baC aþ1ðð ÞÞC bþ1aððdcÞbÞ, we get the desired equality(51).
Corollary 14. If we takea ¼ b ¼ 1 inLemma13, we get
f aþ b 2 ; cþ d 2 1 b a ð Þ Z b a f x;cþ d 2 dx ðd c1 Þ Z d c f aþ b 2 ; y dyþ 1 b a ð Þ d cð Þ Z b a Z d c fðx; yÞdydx ¼ 1 16 bð aÞ d cð Þ Z b a Z d c Z aþb 2 t Z cþd 2 s frsðr; sÞdsdr ( ) dsdt: ð58Þ Theorem 15. Let f: D # R2! R be a partial differentiable
mapping on D :¼ ½a; b ½c; d in R2 with a< b and c < d. If
frs2 L1ðDÞ, then the following equality holds:
f aþ b 2 ; cþ d 2 C b þ 1ð Þ 2 dð cÞb J b cþf aþ b 2 ;d þ Jb df aþ b 2 ;c C a þ 1ð Þ 2 bð aÞa J a bf a; cþ d 2 þ Ja aþf b; cþ d 2 þC a þ1ð ÞC b þ 1ð Þ 4 bð aÞaðd cÞb J a;b aþ;cþfðb;dÞ þ J a;b aþ;dfðb;cÞ þ J a;b b;cþfða;dÞ h þJa;b b;dfða;cÞ i 6jj jfrsj1ðb aÞ d cð Þ 4 21aþ a 1ð Þ a þ 1 21bþ b 1ð Þ b þ 1 " # : ð59Þ
Proof. InLemma 13, taking the modulus, it follows that f aþ b 2 ; cþ d 2 C b þ 1ð Þ 2 dð cÞb J b cþf aþ b 2 ;d þ Jb df aþ b 2 ;c C a þ 1ð Þ 2 bð aÞa J a bf a; cþ d 2 þJa aþf b; cþ d 2 þC a þ 1ð ÞC b þ 1ð Þ 4 bð aÞaðd cÞb J a;b aþ;cþfðb;dÞ þ Ja;baþ;dfðb;cÞ h þJa;b
b;cþfða;dÞ þ Ja;bb;dfða;cÞi ð60Þ
6 ab fjj jrsj1 4 bð aÞaðd cÞb Z b a Z d c b t ð Þa1þ t að Þa1 h i n d shð Þb1þ s cð Þb1i aþ b 2 t cþ d2 s dsdt ¼jj jfrsj1ðb aÞ d cð Þ 4 21aþ a 1ð Þ a þ 1 21bþ b 1ð Þ b þ 1 " # ð61Þ for frs2 L1ðDÞ.
Remark 16. If we takea ¼ b ¼ 1 inTheorem 15, we get
f aþ b 2 ; cþ d 2 1 b a ð Þ Zb a f x;cþ d 2 dx 1 d c ð Þ Zd c f aþ b 2 ; y dy þ 1 b a ð Þ d cð Þ Zb a Z d c fðx; yÞdydx 6jj jfrsj1 16 ðb aÞ d cð Þ: ð62Þ
which is proved by Sarikaya inSarikaya (2012).
Theorem 17. Let f: D # R2! R be a partial differentiable
mapping on D :¼ ½a; b ½c; d in R2 with a< b and c < d. If
frs
j j is a convex function on the co-ordinates on D, then the following equality holds:
f aþ b 2 ; cþ d 2 C b þ 1ð Þ 2 dð cÞb J b cþf aþ b 2 ;d þ Jb df aþ b 2 ;c C a þ 1ð Þ 2 bð aÞa J a bf a; cþ d 2 þ Ja aþf b; cþ d 2 þC a þ 1ð ÞC b þ 1ð Þ 4 bð aÞaðd cÞb J a;b aþ;cþfðb;dÞ þ Ja;baþ;dfðb;cÞ h þJa;b b;cþfða;dÞ þ J a;b b;dfða;cÞi ð63Þ 6 b að Þ d cð Þa2a a þ 1ð Þ2a1þ 1 2aða þ 1Þ b2b b þ 1ð Þ2b1þ 1 2bðb þ 1Þ jfrsða; cÞj þ fjrsða; dÞj þ fjrsðb; cÞj þ fjrsðb; dÞj 4 : ð64Þ
Proof. Since fjrsðr; sÞj is co-ordinates on D, we know that t2 a; b½ ; s 2 c; d½ frsðr;sÞ j j ¼ frs b r b aaþ r a b ab; d s d ccþ s c d cd 6b r b a d s d cjfrsða;cÞj þ b r b a s c d cjfrsða;dÞj þr a b a d s d cjfrsðb;cÞj þ r a b a s c d cjfrsðb;dÞj: ð65Þ From Lemma 13, using co-ordinated convexity of fj j, wers have f aþ b 2 ; cþ d 2 C b þ 1ð Þ 2 dð cÞb J b cþf aþ b 2 ;d þ Jb df aþ b 2 ;c C a þ 1ð Þ 2 bð aÞa J a bf a; cþ d 2 þ Ja aþf b; cþ d 2 þC a þ 1ð ÞC b þ 1ð Þ 4 bð aÞaðd cÞb J a;b aþ;cþfðb;dÞ þ Ja;baþ;dfðb;cÞ h þJa;b
b;cþfða;dÞ þ Ja;bb;dfða;cÞi ð66Þ
6 ab 4 b að Þaðd cÞb Z b a Z d c b t ð Þa1þ t að Þa1 h i n d shð Þb1þ s cð Þb1iZ aþb 2 t Z cþd 2 s frsðr;sÞ j jdsdr ) dsdt ð67Þ 6 ab 4 bð aÞaþ1ðd cÞbþ1 Z b a Z d c b t ð Þa1þ t að Þa1 h i n d shð Þb1þ s cð Þb1iZ aþb 2 t Z cþd 2 s b r ð Þ d sð Þ fjrsða;cÞj ½ þ b rð Þ s cð Þ fjrsða;dÞjþ r að Þ d sð Þ fjrsðb;cÞj þ _ðr aÞ s cð Þ fjrsðb;dÞj i dsdr o dsdt¼ K1þ K2þ K3þ K4: ð68Þ
With a simple calculation, we have K1¼ ab 4 bð aÞaþ1ðd cÞbþ1 Z b a Z d c b t ð Þa1þ t að Þa1 h i d shð Þb1þ s cð Þb1i fjrsða;cÞj Z aþb 2 t Z cþd 2 s b r ð Þ d sð Þdsdr dsdt¼ ab fjrsða;cÞj 4 bð aÞaþ1ðd cÞbþ1 Z b a b t ð Þa1þ t að Þa1 h i Z aþb 2 t b r ð Þdr dt " # Z d c d s ð Þb1þ s cð Þb1 h i Z cþd 2 s d s ð Þds ds " # ð69Þ ¼ ab fjrsða;cÞj 4 bð aÞaþ1ðd cÞbþ1 Z aþb 2 a bt ð Þa1þ t að Þa1 h iZ aþb 2 t br ð Þdrdt " þ Zb aþb 2 b t ð Þa1þ t að Þa1 h iZt aþb 2 b r ð Þdrdt # Zcþd 2 c d s ð Þb1þ s cð Þb1 h iZ cþd 2 s d s ð Þdsds " þ Zd cþd 2 d s ð Þb1þ s cð Þb1 h iZ s cþd 2 d s ð Þdsds # ¼jfrsða;cÞj 4 a2a a þ 1ð Þ2a1þ 1 2aða þ1Þ b2b b þ 1ð Þ2b1þ 1 2bðb þ1Þ ðb aÞ d cð Þ: ð70Þ
Similarly, we also have the following equalities K2¼ ab 4 bð aÞaþ1ðd cÞbþ1 Zb a Zd c b t ð Þa1þ t að Þa1 h i d shð Þb1þ s cð Þb1i f rsða;dÞ j j Zaþb 2 t Zcþd 2 s b r ð Þ s cð Þdsdr dsdt ¼jfrsða;dÞj 4 a2a a þ 1ð Þ2a1þ 1 2aða þ 1Þ b2b b þ 1ð Þ2b1þ 1 2bðb þ 1Þ ðb aÞ d cð Þ; ð71Þ K3¼ ab 4 bð aÞaþ1ðd cÞbþ1 Zb a Zd c b t ð Þa1þ t að Þa1 h i d s ð Þb1þ s cð Þb1 h i fjrsðb;cÞj Zaþb 2 t Zcþd 2 s r a ð Þ d sð Þdsdr dsdt ¼jfrsðb;cÞj 4 a2a a þ 1ð Þ2a1þ 1 2aða þ 1Þ b2b b þ 1ð Þ2b1þ 1 2bðb þ 1Þ ðb aÞ d cð Þ ð72Þ and K4¼ ab 4 bð aÞaþ1ðd cÞbþ1 Zb a Zd c b t ð Þa1þ t að Þa1 h i d shð Þb1þ s cð Þb1i f rsðb;dÞ j jZ aþb 2 t Zcþd 2 s r a ð Þ s cð Þdsdr dsdt ¼jfrsðb;dÞj 4 a2a a þ 1ð Þ2a1þ 1 2aða þ 1Þ b2b b þ 1ð Þ2b1þ 1 2bðb þ 1Þ ðb aÞ d cð Þ: ð73Þ
Thus, if we put the last four equalities in(68), we obtain(64). This completes the proof of the theorem.
Corollary 18. If we takea ¼ b ¼ 1 inTheorem17, we get f aþ b 2 ; cþ d 2 ðb a1 Þ Z b a f x;cþ d 2 dx ðd c1 Þ Zd c f aþ b 2 ;y dy þ 1 b a ð Þ d cð Þ Z b a Zd c fðx;yÞdydx 6ðb aÞ d cð Þ 16 jfrsða;cÞj þ fjrsða;dÞj þ fjrsðb;cÞj þ fjrsðb;dÞj 4 : ð74Þ 3. Conclusion
In this work we give two identities for functions of two variables and apply them to give new Hermite–Hadamard type Fractional integral inequalities for double Fractional integrals involving functions whose derivatives are bounded or co-ordinates convex function onD :¼ ½a; b ½c; d in R2 with
a< b; c < d. Acknowledgement
The authors would like to express their appreciation to the referees for their valuable suggestions which helped to better presentation of this paper.
References
Alomari, M., Darus, M., Dragomir, S.S., 2009. New inequalities of Hermite–Hadamard type for functions whose second derivates absolute values are quasi-convex, RGMIA Res. Rep. Coll., 12 Supplement, Article 14.
Azpeitia, A.G., 1994. Convex functions and the Hadamard inequality. Rev. Colomb. Math. 28, 7–12.
Bakula, M.K., Pecˇaric´, J., 2004. Note on some Hadamard-type inequalities. J. Inequal. Pure Appl. Math. 5 (3), article 74.
Dragomir, S.S., 2001. On Hadamard’s inequality for convex functions on the co-ordinates in a rectangle from the plane. Taiwanese J. Math. 4, 775–788.
Dragomir, S.S., Pearce, C.E.M., 2000. Selected Topics on Hermite– Hadamard Inequalities and Applications. RGMIA Monographs, Victoria University.
Gorenflo, R., Mainardi, F., 1997. Fractional Calculus: Integral and Differential Equations of Fractional Order. Springer Verlag, Wien, 223–276.
Hussain, S., Bhatti, M.I., Iqbal, M., 2009. Hadamard-type inequalities for s-convex functions I. Punjab Univ. J. Math. 41, 51–60.
Kilbas, A.A., Srivastava, H.M., Trujillo, J.J., 2006. Theory and Applications of Fractional Differential Equations, North-Holland Mathematics Studies, 204. Elsevier Sci. B.V., Amsterdam.
Kırmacı, U.S., Dikici, R., 2013. On some Hermite–Hadamard type inequalities for twice differentiable mappings and applications. Tamkang J. Math. 44 (1), 41–51.
Latif, M.A., Dragomir, S.S., 2012. On some new inequalities for differentiable co-ordinated convex functions. J. Inequal. Appl. 1 (December), 1–13.
Miller, S., Ross, B., 1993. An Introduction to the Fractional Calculus and Fractional Differential Equations. John Wiley & Sons, USA, p. 2.
Ozdemir, M.E., Avc, M., Set, E., 2010. On some inequalities of Hermite–Hadamard type via m-convexity. Appl. Math. Lett. 23, 1065–1070.
Ozdemir, M.E., Set, E., Sarikaya, M.Z., 2011. Some new Hadamard’s type inequalities for co-ordinated m-convex and ða; mÞ-convex functions. Hacettepe J. Math. Stat. 40, 219–229.
Samko, S.G., Kilbas, A.A., Marichev, O.I., 1993. Fractional Integrals and Derivatives, Theory and Applications. Gordon and Breach, Yverdon et alibi.
Sarikaya, M.Z., 2012. On the Ostrowski type integral inequality for double integrals. Demon. Math. 45, 533–540.
Sarikaya, M.Z., 2014. On the Hermite–Hadamard-type inequalities for co-ordinated convex function via fractional integrals. Integr. Transf. Spec. Funct. 25, 134–147.
Sarikaya, M.Z., 2015. Some inequalities for differentiable co-ordinated convex mappings. Asian-Eur. J. Math. 8 (21 pages).
Sarikaya, M.Z., Aktan, N., 2011. On the generalization of some integral inequalities and their applications. Math. Comput. Model. 54, 2175–2182.
Sarikaya, M.Z., Yaldiz, H., 2013. On the Hadamard’s type inequalities for L-Lipschitzian mapping. Konuralp J. Math. 1, 33–40.
Sarikaya, M.Z., Set, E., Ozdemir, M.E., Dragomir, S.S., 2012. New some Hadamard’s type inequalities for co-ordinated convex func-tions. Tamsui Oxford J. Inf. Math. Sci. 28, 137–152.
Sarikaya, M.Z., Set, E., Yaldiz, H., Basak, N., 2012. Hermite– Hadamard’s inequalities for fractional integrals and related frac-tional inequalities. Math. Comput. Model. 57, 2403–2407.
Sarikaya, M.Z., Set, E., Ozdemir, M.E., 2014a. On some integral inequalities for twice differantiable mappings. Stud. Univ. Babes-Bolyai Math. 59, 11–24.
Sarikaya, M.Z., Budak, H., Yaldiz, H., 2014b. Cebysev type inequal-ities for co-ordinated convex functions. Pure Appl. Math. Lett. 2, 36–40.
Sarikaya, M.Z., Budak, H., Yaldiz, H., 2014c. Some new Ostrowski type inequalities for co-ordinated convex functions. Turkish J. Anal. Numb. Theory 2, 176–182.
Set, E., Sarikaya, M.Z., Ahmad, A., 2011. A generalization of Chebychev type inequalities for first differentiable mappings. Miskolc Math. Notes 12, 245–253.