Research article

Optimality conditions associated with new controlled extremization models

  • Received: 13 March 2024 Revised: 14 April 2024 Accepted: 28 April 2024 Published: 20 May 2024
  • MSC : 26B25, 49J20, 90C17, 90C32, 90C46

  • Applying a parametric approach, in this paper we studied a new class of multidimensional extremization models with data uncertainty. Concretely, first, we derived the robust conditions of necessary optimality. Thereafter, we established robust sufficient optimality conditions by employing the various forms of convexity of the considered functionals. In addition, we formulated an illustrative example to validate the theoretical results.

    Citation: Tareq Saeed. Optimality conditions associated with new controlled extremization models[J]. AIMS Mathematics, 2024, 9(7): 17319-17338. doi: 10.3934/math.2024842

    Related Papers:

    [1] Serap Özcan, Saad Ihsan Butt, Sanja Tipurić-Spužević, Bandar Bin Mohsin . Construction of new fractional inequalities via generalized n-fractional polynomial s-type convexity. AIMS Mathematics, 2024, 9(9): 23924-23944. doi: 10.3934/math.20241163
    [2] Hasan Kara, Hüseyin Budak, Mehmet Eyüp Kiriş . On Fejer type inequalities for co-ordinated hyperbolic ρ-convex functions. AIMS Mathematics, 2020, 5(5): 4681-4701. doi: 10.3934/math.2020300
    [3] Wei Gao, Artion Kashuri, Saad Ihsan Butt, Muhammad Tariq, Adnan Aslam, Muhammad Nadeem . New inequalities via n-polynomial harmonically exponential type convex functions. AIMS Mathematics, 2020, 5(6): 6856-6873. doi: 10.3934/math.2020440
    [4] Naila Mehreen, Matloob Anwar . Ostrowski type inequalities via some exponentially convex functions with applications. AIMS Mathematics, 2020, 5(2): 1476-1483. doi: 10.3934/math.2020101
    [5] Tekin Toplu, Mahir Kadakal, İmdat İşcan . On n-Polynomial convexity and some related inequalities. AIMS Mathematics, 2020, 5(2): 1304-1318. doi: 10.3934/math.2020089
    [6] Ghulam Farid, Saira Bano Akbar, Shafiq Ur Rehman, Josip Pečarić . Boundedness of fractional integral operators containing Mittag-Leffler functions via (s,m)-convexity. AIMS Mathematics, 2020, 5(2): 966-978. doi: 10.3934/math.2020067
    [7] M. Emin Özdemir, Saad I. Butt, Bahtiyar Bayraktar, Jamshed Nasir . Several integral inequalities for (α, s,m)-convex functions. AIMS Mathematics, 2020, 5(4): 3906-3921. doi: 10.3934/math.2020253
    [8] Anjum Mustafa Khan Abbasi, Matloob Anwar . Ostrowski type inequalities for exponentially s-convex functions on time scale. AIMS Mathematics, 2022, 7(3): 4700-4710. doi: 10.3934/math.2022261
    [9] Saad Ihsan Butt, Ahmet Ocak Akdemir, Muhammad Nadeem, Nabil Mlaiki, İşcan İmdat, Thabet Abdeljawad . (m,n)-Harmonically polynomial convex functions and some Hadamard type inequalities on the co-ordinates. AIMS Mathematics, 2021, 6(5): 4677-4690. doi: 10.3934/math.2021275
    [10] Muhammad Uzair Awan, Nousheen Akhtar, Artion Kashuri, Muhammad Aslam Noor, Yu-Ming Chu . 2D approximately reciprocal ρ-convex functions and associated integral inequalities. AIMS Mathematics, 2020, 5(5): 4662-4680. doi: 10.3934/math.2020299
  • Applying a parametric approach, in this paper we studied a new class of multidimensional extremization models with data uncertainty. Concretely, first, we derived the robust conditions of necessary optimality. Thereafter, we established robust sufficient optimality conditions by employing the various forms of convexity of the considered functionals. In addition, we formulated an illustrative example to validate the theoretical results.



    Calculus is a branch of mathematics which helps us to study the derivatives and integrals. The classical derivative was convoluted with the strength regulation kind kernel and eventually, this gave upward thrust to new calculus referred to as the quantum calculus. Quantum calculus (named q-calculus) is the study of calculus without limits. In recent decades, the quantum calculus has become a powerful tool in numerous branches of mathematics and physics like q-calculus, particularly q-fractional calculus, q-integral calculus, q-transform analysis. Jackson [1] is the first researcher to define the q-analogue of derivative and integral operator as well as provided its applications. It is imperative to mention that quantum integral inequalities are more practical and informative than their classical counterparts. It has been mainly due to the fact that quantum integral inequalities can describe the hereditary properties of the processes and phenomena under investigation. Historically the subject of quantum calculus can be traced back to Euler and Jacobi, but in recent decades it has experienced a rapid development, see [2,3,4,5,6,7]. As a result, new generalizations of the classical concepts of quantum calculus have been initiated and reviewed in many literature. Tariboon and Ntouyas [8,9] proposed the quantum calculus concepts on finite intervals and obtained several q-analogues of classical mathematical objects. This inspired other researchers to establish numerous novel results concerning quantum analogs of classical mathematical results. Noor et al. [10] provided the q-analogues of many known inequalities via the first order q-differentiable convex functions. Humaira et al. [11] obtained a new generalized q1q2-integral identity and established several new q1q2-analogues of first order q1q2-differentiable convex functions over finite rectangles. Wu et al. [12] gave a new corrected q-analogue of the classical Simpson inequality for preinvex function. Deng et al. [13] obtained a new generalized q-integral identity and found several new q-analogues for twice q-differentiable generalized (s,m)-preinvex functions.

    The theory of post quantum calculus denoted by (p,q)-calculus is a natural generalization of the quantum calculus denoted by q-calculus, which has been studied extensively. Recently, Tunç and Göv [14] studied the concept of (p,q)-calculus on the intervals [χ1,χ2]R, defined the (p,q)-derivative and (p,q)-integral and established their basic properties and integral inequalities. Integral inequalities [15,16,17,18,19,20,21,22,23,24,25,26,27] play an important role in understanding the universe, and they can be used to find the uniqueness and existence of the linear and nonlinear differential equations. While convexity is an indispensable tool in the study of inequality theory [28,29,30,31,32,33,34,35,36,37,38].

    It is well-known that the Hermite-Hadamard inequality [39,40,41,42,43] is one of the most important inequalities in the convex functions theory, which can be stated as follows.

    Let K:IRR be a convex function. Then the double inequality

    K(χ1+χ22)1χ2χ1χ2χ1K(t)dtK(χ1)+K(χ2)2 (1.1)

    holds for all χ1,χ2I with χ1χ2.

    As the generalization and refinement of the Hermite-Hadamard inequality (1.1), the Ostrowski inequality [44] can be stated in Theorem 1.1.

    Theorem 1.1. Let K:[χ1,χ2]RR be continuous on [χ1,χ2] and differentiable on (χ1,χ2) such that |K(τ)|M for all τ(χ1,χ2). Then the inequality

    |K(ϱ)1χ2χ1χ2χ1K(τ)dτ|[14+(ϱχ1+χ22)2(χ2χ1)2](χ2χ1)M (1.2)

    holds for all ϱ[χ1,χ2] with the best possible constant 14.

    Inequality (1.2) can be rewritten in its equivalent form

    |K(ϱ)1χ2χ1χ2χ1K(τ)dτ|Mχ2χ1[(ϱχ1)2+(χ2ϱ)22].

    Recently, the generalizations, variants and applications of the Ostrowski inequality have attracted the attention of many researchers.

    Now, we discuss some connections between the class of convex functions and s-type convex functions.

    Definition 1.2. Let s[0,1]. Then the function K:IR is said to be a s-type convex function on I if the inequality

    K(χϱ+(1χ)ρ)[1s(1χ)]K(ϱ)+[1sχ]K(ρ) (1.3)

    holds for all ϱ,ρI and χ[0,1].

    Remark 1. Definition 1.2 leads to the conclusion that

    (1) If we choose s=1, then we get classical convex function.

    (2) If we take s=0, then we have the definition of P-function [45].

    (3) If K is s-type convex function on I, then the range of K is [0,).

    Indeed, let ϱI. Then it follows from the s-type convexity of K that

    K(χη1+(1χ)ϱ)[1s(1χ)]K(η1)+[1sχ]K(ϱ)

    for all η1I and χ[0,1]. If we choose χ=1, then we get

    K(η1)K(η1)+(1s)K(ϱ)
    (1s)K(ρ)0K(ϱ)0

    for all ϱI.

    Proposition 1. Every non-negative convex function is also s-type convex function.

    Proof. The proof is clearly due to

    s(1χ)(1χ),χsχ

    for all χ[0,1] and s[0,1].

    Next, We recall the definition of n-polynomial s-type convex function.

    Definition 1.3. Let s[0,1] and nN. Then the function K:IR is said to be a n-polynomial s-type convex function on I if the inequality

    K(χϱ+(1χ)ρ)1nni=1[1(s(1χ))i]K(ϱ)+1nni=1[1(sχ)i]K(ρ) (1.4)

    holds for ϱ,ρI and χ[0,1].

    Remark 2. From Definition 1.3, one has

    (1) If we choose s=0, then we get the P-function [45].

    (2) If we take s=1, then we obtain the Definition given in [46].

    (3) If we choose n=1, then we obtain Definition 1.2.

    (4) If K is a n-polynomial s-type convex function, then the range of K is [0,).

    Remark 3. Every non-negative n-polynomial convex function is also n-polynomial s-type convex function. Indeed

    1nni=1[1(s(1χ))i]1nni=1[1(1χ)i]

    and

    1nni=1[1χi]1nni=1[1(sχ)i]

    for all χ[0,1],nN and s[0,1].

    In what follows, we suppose that χ1,χ2,χ3,χ4R with χ1<χ2 and χ3<χ4, 0<qk<pk1(k=1,2) are constants, N=[χ1,χ2]×[χ3,χ4]R2 is a rectangle and N=(χ1,χ2)×(χ3,χ4) is the interior of N.

    Definition 1.4. Let K:NR be a continuous function. Then the partial (p1,q1)-, (p2,q2)- and (p1p2,q1q2)-derivatives at (z,w)N are respectively defined by

    χ1p1,q1K(z,w)χ1p1,q1z=K(p1z+(1p1)χ1,w)K(q1z+(1q1)χ1,w)(p1q1)(zχ1)(zχ1),χ3p2,q2K(z,w)χ3p2,q2w=K(z,p2w+(1p2)χ3)K(z,q2w+(1q2)χ3)(p2q2)(wχ3)(wχ3),χ1,χ32p1p2,q1q2K(z,w)χ1p1,q1zχ3p2,q2w=1(p1q1)(p2q2)(zχ1)(wχ3)[K(q1z+(1q1)χ1,q2w+(1q2)χ3)K(q1z+(1q1)χ1,p2w+(1p2)χ3)K(p1z+(1p1)χ1,q2w+(1q2)χ3)+K(p1z+(1p1)χ1,p2w+(1p2)χ3)].

    Definition 1.5. Let K:NR be a continuous function. Then the definite (p1p2,q1q2)-integral on N is defined by

    tχ3sχ1K(z,w)χ1dp1,q1zχ3dp2,q2w=(p1q1)(p2q2)(sχ1)(tχ3)×m=0n=0qn1qm2pn+11pm+12K(qn1pn+11s+(1qn1pn+11)χ1,qm2pm+12t+(1qm2pm+12)χ3) (1.5)

    for (s,t)N.

    Definition 1.6. For any real number n, the (p,q)-analogue is defined by

    [n]p,q=pnqnpq,

    where 0<q<p1 are constants.

    In the present paper, we introduce new Definitions 1.7 and 1.8 for (p1p2,q1q2)-differentiable function, and (p,q)χ4χ1,(p,q)χ2χ3 and (p,q)χ2,χ4 integrals for two variables mappings over finite rectangles by using convex set. These new definitions will open new doors for convexity and (p,q)-calculus for two variables functions over the finite rectangles in the plane R2. We drive a key lemma for (p1p2,q1q2)-integral. By making use of new identity, we will obtain estimates of integral inequality whose twice partial (p1p2,q1q2)-derivatives in absolute value at certain powers are n-polynomial s-type convex function. We also discuss some new special cases of the main results. A briefly conclusion is given at the end.

    Definition 1.7. Let K:NR be a continuous function. Then the partial (p1,q1)-, (p2,q2)- and (p1p2,q1q2)-derivatives at (z,w)N are respectively defined by

    χ2p1,q1K(z,w)χ2p1,q1z=K(p1z+(1p1)χ2,w)K(q1z+(1q1)χ2,w)(p1q1)(χ2z)(χ2z),χ4p2,q2K(z,w)χ4p2,q2w=K(z,p2w+(1p2)χ4)K(z,q2w+(1q2)χ4)(p2q2)(χ4w)(χ4w),χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w=1(p1q1)(p2q2)(zχ1)(χ4w)[K(q1z+(1q1)χ1,q2w+(1q2)χ4)K(q1z+(1q1)χ1,p2w+(1p2)χ4)K(p1z+(1p1)χ1,q2w+(1q2)χ4)+K(p1z+(1p1)χ1,p2w+(1p2)χ4)](zχ1,χ4w),χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w=1(p1q1)(p2q2)(χ2z)(wχ3)[K(q1z+(1q1)χ2,q2w+(1q2)χ3)K(q1z+(1q1)χ2,p2w+(1p2)χ3)K(p1z+(1p1)χ2,q2w+(1q2)χ3)+K(p1z+(1p1)χ2,p2w+(1p2)χ3)](χ2z,wχ3),χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w=(p1q1)(p2q2)(χ2z)(χ4w)[K(q1z+(1q1)χ2,q2w+(1q2)χ4)K(q1z+(1q1)χ2,p2w+(1p2)χ4)K(p1z+(1p1)χ2,q2w+(1q2)χ4)+K(p1z+(1p1)χ2,p2w+(1p2)χ4)](χ2z,χ4w).

    Definition 1.8. Let K:NR be a continuous function. Then the definite (p,q)χ4χ1,(p,q)χ2χ3 and (p,q)χ2,χ4 integrals on [χ1,χ2]×[χ3,χ4] are respectively defined by

    χ4tsχ1K(z,w)χ1dp1,q1zχ3dp2,q2w=(p1q1)(p2q2)(sχ1)(χ4t)×m=0n=0qn1qm2pn+11pm+12K(qn1pn+11s+(1qn1pn+11)χ1,qm2pm+12t+(1qm2pm+12)χ4),
    tχ3χ2sK(z,w)χ2dp1,q1zχ3dp2,q2w(p1q1)(p2q2)(χ2s)(tχ3)×m=0n=0qn1qm2pn+11pm+12K(qn1pn+11s+(1qn1pn+11)χ2,qm2pm+12t+(1qm2pm+12)χ3)

    and

    χ4tχ2sK(z,w)χ2dp1,q1zχ4dp2,q2w(p1q1)(p2q2)(χ2s)(χ4t)×m=0n=0qn1qm2pn+11pm+12K(qn1pn+11s+(1qn1pn+11)χ2,qm2pm+12t+(1qm2pm+12)χ4)

    for (s,t)[χ1,χ2]×[χ3,χ4].

    Lemma 2.1. Let K:NR be a twice partial (p1p2,q1q2)-differentiable function on No such that the partial (p1p2,q1q2)-derivatives χ1,χ32p1p2,q1q2K(z,w)χ1p1,q1zχ3p2,q2w, χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w, χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w and χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w are continuous and integrable on N. Then one has the equality

    K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2v+χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2v]+A=Δ{(ϱχ1)2(ρχ3)21010zwχ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w0dp1,q1z0dp2,q2w+(ϱχ1)2(χ4ρ)21010zwχ4χ12p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ4)χ1p1,q1zχ4p2,q2w0dp1,q1z0dp2,q2w+(χ2ϱ)2(ρχ3)21010zwχ2χ32p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ3)χ3p1,q1zχ2p2,q2w0dp1,q1z0dp2,q2w+(χ2ϱ)2(χ4ρ)21010zwχ2,χ42p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ4)χ2p1,q1zχ4p2,q2w0dp1,q1z0dp2,q2w}, (2.1)

    where

    A=1p1p2(χ2χ1)(χ4χ3)p1ϱ+(1p1)χ1χ1p2ρ+(1p2)χ3χ3K(u,v)χ1dp1,q1uχ3dp2,q2v+1p1p2(χ2χ1)(χ4χ3)p1ϱ+(1p1)χ1χ1χ4p2ρ+(1p2)χ4K(u,v)χ1dp1,q1uχ4dp2,q2v+1p1p2(χ2χ1)(χ4χ3)χ2p1ϱ+(1p1)χ2p2ρ+(1p2)χ3χ3K(u,v)χ2dp1,q1uχ3dp2,q2v+1p1p2(χ2χ1)(χ4χ3)χ2p1ϱ+(1p1)χ2χ4p2ρ+(1p2)χ4K(u,v)χ2dp1,q1uχ4dp2,q2v

    for all ϱ,ρN and Δ=q1q2(χ2χ1)(χ4χ3).

    Proof. We consider the integral

    1010zwχ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w0dp1,q1z0dp2,q2w.

    By the definition of partial (p1p2,q1q2)-derivative and definite (p1p2,q1q2)χ1,χ3-integral, we have

    1010zwχ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w0dp1,q1z0dp2,q2w=1(1q1)(1q2)(ϱχ1)(ρχ3)×[1010K(zq1ϱ+(1zq1)χ1,wq2ρ+(1wq2)χ3)0dp1,q1z0dp2,q2w1010K(zq1ϱ+(1zq1)χ1,wp2ρ+(1wp2)χ3)0dp1,q1z0dp2,q2w1010K(zp1ϱ+(1zp1)χ1,wq2ρ+(1wq2)χ3)0dp1,q1z0dp2,q2w+1010K(zp1ϱ+(1zp1)χ1,wp2ρ+(1wp2)χ3)0dp1,q1z0dp2,q2w]=1(ϱχ1)(ρχ3)×[n=0m=0qn1qm2pn+11pm+12K(qn+11pn+11ϱ+(1qn+11pn+11)χ1,qm+12pm+12ρ+(1qm+12pm+12)χ3)n=0m=0qn1qm2pn+11pm+12K(qn+11pn+11ϱ+(1qn+11pn+11)χ1,qm2pm2ρ+(1qm2pm2)χ3)n=0m=0qn1qm2pn+11pm+12K(qn1pn+11ϱ+(1qn1pn+11)χ1,qm+12pm+12ρ+(1qm+12pm+12)χ3)+n=0m=0qn1qm2pn+11pm+12K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)]=1q1q2(ϱχ1)(ρχ3)n=1m=1qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)1p2q1(ϱχ1)(ρχ3)n=1m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)1p1q2(ϱχ1)(ρχ3)n=0m=1qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)+1p1p2(ϱχ1)(ρχ3)n=0m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3). (2.2)

    Note that

    1q1q2(ϱχ1)(ρχ3)n=1m=1qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)=K(ϱ,ρ)q1q2(ϱχ1)(ρχ3)1q1q2(ϱχ1)(ρχ3)n=0qn1pn1K(qn1pn1ϱ+(1qn1pn1)χ1,ρ)1q1q2(ϱχ1)(ρχ3)m=0qm2pm2K(ϱ,qm2pm2ρ+(1qm2pm2)χ3)+1q1q2(ϱχ1)(ρχ3)×n=0m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3), (2.3)
    1p2q1(ϱχ1)(ρχ3)n=1m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)=1p2q1(ϱχ1)(ρχ3)m=0qm2pm2K(ϱ,qm2pm2ρ+(1qm2pm2)χ3)1p2q1(ϱχ1)(ρχ3)×n=0m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3) (2.4)

    and

    1p1q2(ϱχ1)(ρχ3)n=0m=1qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)=1p1q2(ϱχ1)(ρχ3)n=0qn1pn1K(qn1pn1ϱ+(1qn1pn1)χ1,ρ)1p1q2(ϱχ1)(ρχ3)×n=0m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3). (2.5)

    Utilizing (2.3)–(2.5) in (2.2), we get

    1010zwχ1,χ32p1p2,q1q2χ1p1,q1zχ3p2,q2wK(zϱ+(1z)χ1,wρ+(1w)χ3)0dp1,q1z0dp2,q2w=K(ϱ,ρ)q1q2(ϱχ1)(ρχ3)(p2q2)(ρχ3)p2q1q2(ϱχ1)(ρχ3)2m=0qm2pm2K(ϱ,qm2pm2ρ+(1qm2pm2)χ3)(p1q1)(ϱχ1)p1q1q2(ϱχ1)2(ρχ3)n=0qn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,ρ)+(p1q1)(p2q2)(ϱχ1)(ρχ3)p1p2q1q2(ϱχ1)2(ρχ3)2n=0m=0qn1qm2pn1pm2K(qn1pn1ϱ+(1qn1pn1)χ1,qm2pm2ρ+(1qm2pm2)χ3)

    and

    1010zwχ1,χ32p1p2,q1q2χ1p1,q1zχ3p2,q2wK(zϱ+(1z)χ1,wρ+(1w)χ3)0dp1,q1z0dp2,q2w=K(ϱ,ρ)q1q2(ϱχ1)(ρχ3)1p2q1q2(ϱχ1)(ρχ3)2p2ρ+(1p2)χ3χ3K(ϱ,v)0dp2,q2v1p1q1q2(ϱχ1)2(ρχ3)p1ϱ+(1p1)χ1χ1K(u,ρ)0dp1,q1u+1p1p2q1q2(ϱχ1)2(ρχ3)2p1ϱ+(1p1)χ1χ1p2ρ+(1p2)χ3χ3K(u,v)0dp1,q1u0dp2,q2v. (2.6)

    Multiplying both sides of equality (2.6) by Δ(ϱχ1)2(ρχ3)2 leads to

    Δ(ϱχ1)2(ρχ3)21010zwχ1,χ32p1p2,q1q2χ1p1,q1zχ3p2,q2wK(zϱ+(1z)χ1,wρ+(1w)χ3)0dp1,q1z0dp2,q2w=(ϱχ1)(ρχ3)(χ2χ1)(χ4χ3)K(ϱ,ρ)ϱχ1p2(χ2χ1)(χ4χ3)p2ρ+(1p2)χ3χ3K(ϱ,v)0dp2,q2vρχ3p1(χ2χ1)(χ4χ3))p1ϱ+(1p1)χ1χ1K(u,ρ)0dp1,q1u+1p1p2(χ2χ1)(χ4χ3)p1ϱ+(1p1)χ1χ1p2ρ+(1p2)χ3χ3K(u,v)0dp1,q1u0dp2,q2v. (2.7)

    Similarly, calculating the remaining integrals and by using definition 1.8 we get

    Δ(ϱχ1)2(χ4ρ)21010zwχ4χ12p1p2,q1q2χ1p1,q1zχ4p2,q2wK(zϱ+(1z)χ1,wρ+(1w)χ4)0dp1,q1z0dp2,q2w=(ϱχ1)(χ4ρ)(χ2χ1)(χ4χ3)K(ϱ,ρ)ϱχ1p2(χ2χ1)(χ4χ3)χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2vχ4ρp1(χ2χ1)(χ4χ3))p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+1p1p2(χ2χ1)(χ4χ3)p1ϱ+(1p1)χ1χ1χ4p2ρ+(1p2)χ4K(u,v)χ1dp1,q1uχ4dp2,q2v, (2.8)
    Δ(χ2ϱ)2(ρχ3)21010zwχ2χ32p1p2,q1q2χ2p1,q1zχ3p2,q2wK(zϱ+(1z)χ1,wρ+(1w)χ3)0dp1,q1z0dp2,q2w=(χ2ϱ)(ρχ3)(χ2χ1)(χ4χ3)K(ϱ,ρ)χ2ϱp2(χ2χ1)(χ4χ3)p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2vρχ3p1(χ2χ1)(χ4χ3))χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u+1p1p2(χ2χ1)(χ4χ3)p1ϱ+(1p1)χ1χ2χ2p2ρ+(1p2)χ2K(u,v)χ2dp1,q1uχ3dp2,q2v, (2.9)
    Δ(χ2ϱ)2(χ4ρ)21010zwχ2,χ42p1p2,q1q2χ2p1,q1zχ4p2,q2wK(zϱ+(1z)χ2,wρ+(1w)χ4)0dp1,q1z0dp2,q2w=(χ2ϱ)(χ4ρ)(χ2χ1)(χ4χ3)K(ϱ,ρ)χ2ϱp2(χ2χ1)(χ4χ3)χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2vχ4ρp1(χ2χ1)(χ4χ3))χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u+1p1p2(χ2χ1)(χ4χ3)χ4p1ϱ+(1p1)χ2χ4p2ρ+(1p2)χ4K(u,v)χ2dp1,q1uχ4dp2,q2v. (2.10)

    From (2.7)–(2.10) and (2.1), we derive the desired result of Lemma 2.1.

    Remark 4. Taking pk=1 for k=1,2 we get

    K(ϱ,ρ)1(χ2χ1)[ϱχ1K(u,ρ) χ1dq1u+χ2ϱK(u,ρ) χ2dq1u]1(χ4χ3)[ρχ3K(ϱ,v) χ3dq2v+χ4ρK(ϱ,v) χ4dq2v]+W=Δ{(ϱχ1)2(ρχ3)21010zw χ1,χ32q1,q2K(zϱ+(1z)χ1,wρ+(1w)χ3) χ1q1z χ3q2w 0dq1z 0dq2w+(ϱχ1)2(χ4ρ)21010zw χ4χ12q1,q2K(zϱ+(1z)χ1,wρ+(1w)χ4) χ1q1z χ4q2w 0dq1z 0dq2w+(χ2ϱ)2(ρχ3)21010zw χ2χ32q1,q2K(zϱ+(1z)χ2,wρ+(1w)χ3) χ3q1z χ2q2w 0dq1z 0dq2w+(χ2ϱ)2(χ4ρ)21010zw χ2,χ42q1,q2K(zϱ+(1z)χ2,wρ+(1w)χ4) χ2q1z χ4q2w 0dq1z 0dq2w}, (2.11)

    where

    W=1(χ2χ1)(χ4χ3)ϱχ1ρχ3K(u,v)χ1dq1uχ3dq2v+1(χ2χ1)(χ4χ3)ϱχ1χ4ρK(u,v)χ1dq1uχ4dq2v+1(χ2χ1)(χ4χ3)χ2ϱρχ3K(u,v)χ2dq1uχ3dq2v+1(χ2χ1)(χ4χ3)χ2ϱχ4ρK(u,v)χ2dq1uχ4dq2v.

    In this section, we introduce (p1p2,q1q2)-Ostrowski inequalities by using n-polynomial s-type convex function on the co-ordinates.

    Theorem 3.1. Suppose that nN, s[0,1] and all the assumptions of Lemma 2.1 are true. If |χ1,χ32p1p2,q1q2K(z,w)χ1p1,q1zχ3p2,q2w|τ2, |χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w|τ2, |χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w|τ2 and |χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w|τ2 are n-polynomial s-type convex functions on the co-ordinates on N for τ1,τ2>1 with 1τ1+1τ2=1 and |χ1,χ32p1p2,q1q2K(ϱ,ρ)χ1p1,q1zχ3p2,q2w|M, |χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w|M, |χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w|M, |χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w|M, ϱ,ρN, then the following inequality holds

    |K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2v+χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2v]+A|ΔMτ2(Cp1,q1+Dp1,q1)(Cp2,q2+Dp2,q2)[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2]τ1[1+τ1]p1,q1[1+τ1]p2,q2, (3.1)

    where

    Cpk,qk=1(pkqk)nni=1e=0siqekpe+1k(1qekpe+1k)i,Dpk,qk=11nni=1si(pkqkpi+1kqi+1k)

    for k=1,2 and Δ,A are defined in Lemma 2.1.

    Proof. Taking absolute value on both sides of (2.1), by applying Hölder inequality for double integrals and utilizing the fact that |2p1p2,q1q2Kp1,q1zp2,q2w|τ2 is n-polynomial s-type convex on co-ordinates, we get the following inequality

    |K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2v+χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2v]+A|Δ(1010zτ1wτ10dp1,q1z0dp2,q2w)1τ1×{(ϱχ1)2(ρχ3)2(1010|χ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ20dp1,q1z0dp2,q2w)1τ2
    +(ϱχ1)2(χ4ρ)2(1010|χ4χ12p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ4)χ1p1,q1zχ4p2,q2w|τ20dp1,q1z0dp2,q2w)1τ2+(χ2ϱ)2(ρχ3)2(1010|χ2χ32p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ3)χ2p1,q1zχ3p2,q2w|τ20dp1,q1z0dp2,q2w)1τ2+(χ2ϱ)2(χ4ρ)2(1010|χ2,χ42p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ4)χ2p1,q1zχ4p2,q2w|τ20dp1,q1z0dp2,q2w)1τ2}.

    Considering first integral

    1010|χ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ20dp1,q1z0dp2,q2w10{10[1nni=1[1(s(1z))i]|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ2+1nni=1[1(sz)i]|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ1p2,q2w|τ2]0dp1,q1z}0dp2,q2w. (3.2)

    Computing the (p1,q1)-integral on the right-hand side of (3.2), we have

    10[1nni=1[1(s(1z))i]|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ2+1nni=1[1(sz)i]|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ1p2,q2w|τ2]0dp1,q1z.

    In view of the Definitions 1.4 for k=1,2, we get

    Cpk,qk=1nni=110[1(s(1z))i]0dpk,qkz=1(pkqk)nni=1e=0siqekpe+1k(1qekpe+1k)i,Dpk,qk=1nni=110[1(sz)i]0dpk,qkz=11nni=1si(pkqkpi+1kqi+1k).

    Putting the above calculations into (3.2), we obtain

    10[Cp1,q1|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ2+Dp1,q1|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ2]0dp2,q2w. (3.3)

    Similarly, by computing the (p2,q2)-integral, utilizing the fact |χ1,χ32p1p2,q1q2K(ϱ,ρ)χ1p1,q1zχ3p2,q2w|M,ϱ,ρN on the right-hand side of (3.3), we have

    1010|χ1,χ32p1p2,q1q2K(zϱ+(1w)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ20dp1,q1z0dp2,q2wMτ2(Cp1,q1+Dp1,q1)(Cp2,q2+Dp2,q2). (3.4)

    Analogously, we get

    1010|χ4χ12p1p2,q1q2K(zϱ+(1w)χ1,χ4+wρ+(1w)χ4)χ1p1,q1zχ4p2,q2w|τ20dp1,q1z0dp2,q2wMτ2(Cp1,q1+Dp1,q1)(Cp2,q2+Dp2,q2) (3.5)
    1010|χ2χ32p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ3)χ2p1,q1zχ3p2,q2w|τ20dp1,q1z0dp2,q2wMτ2(Cp1,q1+Dp1,q1)(Cp2,q2+Dp2,q2) (3.6)

    and

    1010|χ2,χ42p1p2,q1q2K(zϱ+(1z)χ2,χ4+wρ+(1w)χ4)χ2p1,q1zχ4p2,q2w|τ20dp1,q1z0dp2,q2wMτ2(Cp1,q1+Dp1,q1)(Cp2,q2+Dp2,q2). (3.7)

    Now by making use of the inequalities (3.4)(3.7) and using the fact that

    1010zτ1wτ10dp1,q1z0dp2,q2w=1[1+τ1]p1,q1[1+τ1]p2,q2,

    we get the desired inequality (3.1). This completes the proof.

    Corollary 1. I. Taking pk=1 for k=1,2 in Theorem 3.1, we get

    |K(ϱ,ρ)1(χ2χ1)[ϱχ1K(u,ρ)χ1dq1u+χ2ϱK(u,ρ)χ2dq1u]1(χ4χ3)[ρχ3K(ϱ,v)χ3dq2v+χ4ρK(ϱ,v)χ4dq2v]+W|ΔMτ2(Cq1+Dq1)(Cq2+Dq2)[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2]τ1[1+τ1]q1[1+τ1]q2,

    where

    Cqk=1(1qk)nni=1e=0siqek(1qek)i,Dqk=11nni=1si(1qk1qi+1k)

    and Δ,W are defined in Remark 4.

    II. Taking qk1 for k=1,2 in part I, we get

    |K(ϱ,ρ)+1(χ2χ1)(χ4χ3)χ2χ1χ4χ3K(u,v)dvduQ|Mτ2(2nni=1(i+1sii+1))2τ1(1+τ1)2[(ϱχ1)2+(χ2ϱ)2χ2χ1][(ρχ3)2+(χ4ρ)2χ4χ3],

    where

    Q=1χ2χ1χ2χ1K(u,ρ)du+1χ4χ3χ4χ3K(ϱ,v)dv.

    Remark 5. Taking n=s=1 in part II of Corollary 1, we obtain Theorem 4 of [47].

    Theorem 3.2. Suppose that nN, s[0,1] and all the assumptions of Lemma 2.1 holds. If |χ1,χ32p1p2,q1q2K(z,w)χ1p1,q1zχ3p2,q2w|τ, |χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w|τ, |χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w|τ and |χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w|τ are n-polynomial s-type convex functions on the co-ordinates on N for τ1, and |χ1,χ32p1p2,q1q2K(ϱ,ρ)χ1p1,q1zχ3p2,q2w|M, |χ4χ12p1p2,q1q2K(z,w)χ1p1,q1zχ4p2,q2w|M, |χ2χ32p1p2,q1q2K(z,w)χ2p1,q1zχ3p2,q2w|M, |χ2,χ42p1p2,q1q2K(z,w)χ2p1,q1zχ4p2,q2w|M, ϱ,ρN, then the following inequality holds

    |K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2v+χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2v]+A|ΔMτ(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2)[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2][(p1+q1)(p2+q2)]11τ, (3.8)

    where

    Apk,qk=1pk+qk(pkqk)nni=1e=0siq2ekp2e+2k(1qekpe+1k)i,Bpk,qk=1pk+qk1nni=1si(pkqkpi+2kqi+2k)

    for k=1,2 and Δ,A are defined in Lemma 2.1.

    Proof. Taking absolute value on both sides of (2.1), by applying power mean inequality for double integrals, we get the following inequality

    |K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)0dp1,q1u+p1ϱ+(1p1)χ2χ2K(u,ρ)0dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)0dp2,q2v+p2ρ+(1p2)χ4χ4K(ϱ,v)0dp2,q2v]+A|Δ(1010zw0dp1,q1z0dp2,q2w)11τ×{(ϱχ1)2(ρχ3)2(1010zw|χ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ0dp1,q1z0dp2,q2w)1τ
    +(ϱχ1)2(χ4ρ)2(1010zw|χ4χ12p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ4)χ1p1,q1zχ4p2,q2w|τ0dp1,q1z0dp2,q2w)1τ+(χ2ϱ)2(ρχ3)2(1010zw|χ2χ32p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ3)χ2p1,q1zχ3p2,q2w|τ0dp1,q1z0dp2,q2w)1τ+(χ2ϱ)2(χ4ρ)2(1010zw|χ2,χ42p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ4)χ2p1,q1zχ4p2,q2w|τ0dp1,q1z0dp2,q2w)1τ}.

    Considering first integral

    1010zw|χ1,χ32p1p2,q1q2K(zϱ+(1z)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ0dp1,q1z0dp2,q2w10w{10z[1nni=1[1(s(1z))i]|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ+1nni=1[1(sz)i]|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ]0dp1,q1z}0dp2,q2w. (3.9)

    Computing the (p1,q1)-integral on the right-hand side of (3.9), we have

    10z[1nni=1[1(s(1z))i]|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ+1nni=1[1(sz)i]|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ]0dp1,q1z.

    In view of the Definitions 1.5 fpr k=1,2, we get

    Apk,qk=1nni=110z[1(s(1z))i]0dpk,qkz=1pk+qk(pkqk)nni=1e=0siq2ekp2e+2k(1qekpe+1k)i,Bpk,qk=1nni=110z[1(sz)i]0dpk,qkz=1pk+qk1nni=1si(pkqkpi+2kqi+2k).

    Putting the above calculations into (3.9), we obtain

    10w[Ap1,q1|χ1,χ32p1p2,q1q2K(ϱ,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ+Bp1,q1|χ1,χ32p1p2,q1q2K(χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ]0dp2,q2w. (3.10)

    Similarly, by computing the (p2,q2)-integral, utilizing the fact |χ1,χ32p1p2,q1q2K(ϱ,ρ)χ1p1,q1zχ3p2,q2w|M,ϱ,ρN on the right-hand side of (3.10), we have

    1010zw|χ1,χ32p1p2,q1q2K(zϱ+(1w)χ1,wρ+(1w)χ3)χ1p1,q1zχ3p2,q2w|τ0dp1,q1z0dp2,q2wMτ(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2). (3.11)

    Analogously, we get

    1010zw|χ4χ12p1p2,q1q2K(zϱ+(1w)χ1,χ4+wρ+(1w)χ4)χ1p1,q1zχ4p2,q2w|τ0dp1,q1z0dp2,q2wMτ(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2), (3.12)
    1010zw|χ2χ32p1p2,q1q2K(zϱ+(1z)χ2,wρ+(1w)χ3)χ2p1,q1zχ3p2,q2w|τ0dp1,q1z0dp2,q2wMτ(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2) (3.13)

    and

    1010zw|χ2,χ42p1p2,q1q2K(zϱ+(1z)χ2,χ4+wρ+(1w)χ4)χ2p1,q1zχ4p2,q2w|τ0dp1,q1z0dp2,q2wMτ(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2). (3.14)

    Now by making use of the inequalities (3.11)(3.14) and the fact that

    1010zw0dp1,q1z0dp2,q2w=1(p1+q1)(p2+q2),

    we get the desired inequality (3.8). This completes the proof.

    Corollary 2. I. Taking τ=1 in Theorem 3.2, we get

    |K(ϱ,ρ)1p1(χ2χ1)[p1ϱ+(1p1)χ1χ1K(u,ρ)χ1dp1,q1u+χ2p1ϱ+(1p1)χ2K(u,ρ)χ2dp1,q1u]1p2(χ4χ3)[p2ρ+(1p2)χ3χ3K(ϱ,v)χ3dp2,q2v+χ4p2ρ+(1p2)χ4K(ϱ,v)χ4dp2,q2v]+A|ΔM(Ap1,q1+Bp1,q1)(Ap2,q2+Bp2,q2)[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2].

    II. Taking pk=1 for k=1,2 in Theorem 3.2, we obtain

    |K(ϱ,ρ)1(χ2χ1)[ϱχ1K(u,ρ)χ1dq1u+χ2ϱK(u,ρ)χ2dq1u]1(χ4χ3)[ρχ3K(ϱ,v)χ3dq2v+χ4ρK(ϱ,v)χ4dq2v]+W|ΔMτ(Aq1+Bq1)(Aq2+Bq2)[(1+q1)(1+q2)]11τ[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2],

    where

    Aqk=11+qk1qknni=1e=0siq2ek(1qek)i,Bqk=11+qk1nni=1si(1qk1qi+2k),

    and Δ,W are defined in Remark 4.

    III. Taking pk=1 for k=1,2, and qk1 in part II, we get

    |K(ϱ,ρ)+1(χ2χ1)(χ4χ3)χ2χ1χ4χ3K(u,v)dvduQ|M41τ1(1nni=1i2+5i+66si2(i+1)(i+2))2τ[(ϱχ1)2+(χ2ϱ)2χ2χ1][(ρχ3)2+(χ4ρ)2χ4χ3],

    where Q is defined in part II of Corollary 1.

    IV. Taking n=1=s in part III, we have the following inequality

    |K(ϱ,ρ)+1(χ2χ1)(χ4χ3)χ2χ1χ4χ3K(u,v)dvduQ|M4[(ϱχ1)2+(χ2ϱ)2χ2χ1][(ρχ3)2+(χ4ρ)2χ4χ3],

    where Q is defined in part II of Corollary 1.

    Let kR{1,0}, and φ1 and φ2 be two distinct positive real numbers. Then the generalized logarithmic mean Lk(φ1,φ2) is defined by

    Lk(φ1,φ2)=(φk+12φk+11(k+1)(φ2φ1))1k.

    Proposition 2. If m,k>1 and χ1,χ2,χ3,χ4 are positive real numbers such that χ1<χ2 and χ3<χ4, then one has

    |ϱm×ρkρk(χ2χ1)[Lmm(ϱ,χ1)+Lmm(χ2,ϱ)]ϱk(χ4χ3)[Lkk(ρ,χ3)+Lkk(χ4,ρ)]+1(χ2χ1)(χ4χ3)(Lmm(ϱ,χ1)+Lmm(χ2,ϱ))(Lkk( ρ,χ3)+Lkk(χ4,ρ))|M (1+τ1)2τ1[(ϱχ1)2+(χ2ϱ)2χ2χ1][(ρχ3)2+(χ4ρ)2χ4χ3].

    Proof. Let K(ϱ,ρ)=ϱm×ρk for m,k>1. Then, we have

    ϱχ1um×ρkχ1dq1u=ρk[1q11qm+11](ϱm+1χm+11ϱχ1),
    χ2ϱum×ρkχ2dq1u=ρk[1q11qm+11](χm+12ϱm+1χ2ϱ),
    ρχ3ϱm×vkχ3dq2v=ϱm[1q21qk+12](ρk+1χk+13 ρχ3),
    χ4ρϱm×vkχ4dq2v=ϱm[1q21qk+12](χk+14ρk+1χ4ρ),
    ϱχ1ρχ3um×vkχ1dq1uχ3dq2v=[1q11qm+11][1q21qk+12](ϱm+1χm+11ϱχ1)(ρk+1χk+13ρχ3),
    ϱχ1χ4ρum×vkχ1dq1uχ4dq2v=[1q11qm+11][1q21qk+12](ϱm+1χm+11ϱχ1)(χk+14ρk+1χ4ρ),
    χ2ϱρχ3um×vkχ2dq1uχ3dq2v=[1q11qm+11][1q21qk+12](χm+12ϱm+1χ2ϱ)(ρk+1χk+13ρχ3)

    and

    χ2ϱχ4ρum×vkχ2dq1uχ4dq2v=[1q11qm+11][1q21qk+12](χm+12ϱm+1χ2ϱ)(χk+14ρk+1χ4ρ).

    It follows from part I of Corollary 1 that

    |K(ϱ,ρ)ρk(χ2χ1)[1q11qm+11][(ϱm+1χm+11ϱχ1)+(χm+12ϱm+1χ2ϱ)]ϱk(χ4χ3)[1q21qk+12][(ρk+1χk+13 ρχ3)+(χk+14ρk+1χ4ρ)]+Z|ΔMτ2(Cq1+Dq1)(Cq2+Dq2)[(ϱχ1)2+(χ2ϱ)2][(ρχ3)2+(χ4ρ)2]τ1[1+τ1]q1[1+τ1]q2,

    where

    Z=1(χ2χ1)(χ4χ3)[1q11qm+11][1q21qk+12][(ϱm+1χm+11ϱχ1)(ρk+1χk+13ρχ3)+(ϱm+1χm+11ϱχ1)(χk+14ρk+1χ4ρ)+(χm+12ϱm+1χ2ϱ)(ρk+1χk+13ρχ3)+(χm+12ϱm+1χ2ϱ)(χk+14ρk+1χ4ρ)],
    Cqk=1(1qk)nni=1e=0siqek(1qek)i,Dqk=11nni=1si(1qk1qi+1k).

    Remark 6. Applying the same idea as in Proposition 2 and using Theorems 3.1, 3.2 and their corresponding corollaries, and choosing suitable functions, for example K(ϱ,ρ)=ϱm×ρk,m,k>1 and ϱ,ρ>0;K(ϱ,ρ)=1ϱρ,ϱ,ρ>0;K(ϱ,ρ)=eϱ+ρ,ϱ,ρR, and so on, we can obtain other new interesting inequalities for special means. We omit their proofs and the details are left to the interested readers.

    In this paper, we have defined several new partial post quantum derivatives and integrals for the functions with two variables, provided some new generalizations in the frame of a new class of convex functions named n-polynomial s-type convex functions, found a new version (p1p2,q1q2)-Ostrowski type inequality via the class of n-polynomial s-type convex functions on co-ordinates, established a twice partial integral identitity involving (p1p2,q1q2)-differentiable functions, and generalized the Ostrowski type inequality. Our results are the generalizations of many previous known results, and our ideas and approach may lead to a lot of follow-up research.

    The authors would like to thank the anonymous referees for their valuable comments and suggestions, which led to considerable improvement of the article.

    This work was supported by the National Natural Science Foundation of China (Grant Nos. 11971142, 61673169, 11701176, 11871202).

    The authors declare that they have no competing interests.



    [1] T. Antczak, Parametric approach for approximate efficiency of robust multiobjective fractional programming problems, Math. Methods Appl. Sci., 44 (2021), 11211–11230. https://doi.org/10.1002/mma.7482 doi: 10.1002/mma.7482
    [2] T. Antczak, A. Pitea, Parametric approach to multitime multiobjective fractional variational problems under (f, ρ)-convexity, Optimal Control Appl. Methods, 37 (2016), 831–847. http://dx.doi.org/10.1002/oca.2192 doi: 10.1002/oca.2192
    [3] A. Baranwal, A. Jayswal, Preeti, Robust duality for the uncertain multitime control optimization problems, Internat. J. Robust Nonlinear Control, 32 (2022), 5837–5847. https://doi.org/10.1002/rnc.6113 doi: 10.1002/rnc.6113
    [4] A. Beck, A. Ben-Tal, Duality in robust optimization: Primal worst equals dual best, Oper. Res. Lett., 37 (2009), 1–6. https://doi.org/10.1016/j.orl.2008.09.010 doi: 10.1016/j.orl.2008.09.010
    [5] W. Dinkelbach, On nonlinear fractional programming, Manag. Sci., 13 (1967), 492–498.
    [6] Y. Guo, G. Ye, W. Liu, D. Zhao, S. Treanţă, Optimality conditions and duality for a class of generalized convex interval-valued optimization problems, Mathematics, 9 (2021), 2979. https://doi.org/10.3390/math9222979 doi: 10.3390/math9222979
    [7] R. Jagannathan, Duality for nonlinear fractional programs, Zeitschrift fuer Oper. Res., 17 (1973), 1–3. https://doi.org/10.1007/BF01951364 doi: 10.1007/BF01951364
    [8] A. Jayswal, Preeti, M. A. Jiménez, An exact l1 penalty function method for a multitime control optimization problem with data uncertainty, Optim. Control Appl. Methods., 41 (2020), 1705–1717. https://doi.org/10.1002/oca.2634 doi: 10.1002/oca.2634
    [9] A. Jayswal, Preeti, M. A. Jiménez, Robust penalty function method for an uncertain multi-time control optimization problems, J. Math. Anal. Appl., 505 (2022), 125453. https://doi.org/10.1016/j.jmaa.2021.125453 doi: 10.1016/j.jmaa.2021.125453
    [10] V. Jeyakumar, G. Li, G.M. Lee, Robust duality for generalized convex programming problems under data uncertainty, Nonlinear Anal., 75 (2012), 1362–1373. https://doi.org/10.1016/j.na.2011.04.006 doi: 10.1016/j.na.2011.04.006
    [11] G. S. Kim, M. H. Kim, On sufficiency and duality for fractional robust optimization problems involving (V, ρ)-invex function, East Asian Math. J., 32 (2016), 635–639. https://doi.org/10.7858/eamj.2016.043 doi: 10.7858/eamj.2016.043
    [12] M. H. Kim, G. A. Kim, On optimality and duality for generalized fractional robust optimization problems, East Asian Math. J., 31 (2015), 737–742. http://dx.doi.org/10.7858/eamj.2015.054 doi: 10.7858/eamj.2015.054
    [13] M. H. Kim, G. S. Kim, Optimality conditions and duality in fractional robust optimization problems, East Asian Math. J., 31 (2015), 345–349. https://doi.org/10.7858/eamj.2015.025 doi: 10.7858/eamj.2015.025
    [14] Z. Lu, Y. Zhu, Q. Lu, Stability analysis of nonlinear uncertain fractional differential equations with Caputo derivative, Fractals, 29 (2021), 2150057. https://doi.org/10.1142/S0218348X21500572 doi: 10.1142/S0218348X21500572
    [15] S. S. Manesh, M. Saraj, M. Alizadeh, M. Momeni, On robust weakly ϵ-efficient solutions for multi-objective fractional programming problems under data uncertainty, AIMS Mathematics, 7 (2021), 2331–2347. https://doi.org/10.3934/math.2022132 doi: 10.3934/math.2022132
    [16] N. B. Minh, T. T. T. Phuong, Robust equilibrium in transportation networks, Acta Math. Vietnam., 45 (2020), 635–650. https://doi.org/10.1007/s40306-018-00320-3 doi: 10.1007/s40306-018-00320-3
    [17] Ş. Mititelu, Efficiency and duality for multiobjective fractional variational problems with (ρ, b)-quasiinvexity, Yugosl. J. Oper. Res., 19 (2016).
    [18] Ş. Mititelu, S. Treanţă, Efficiency conditions in vector control problems governed by multiple integrals, J. Appl. Math. Comput., 57 (2018), 647–665. https://doi.org/10.1007/s12190-017-1126-z doi: 10.1007/s12190-017-1126-z
    [19] C. Nahak, Duality for multiobjective variational control and multiobjective fractional variational control problems with pseudoinvexity, Int. J. Stoch. Anal., 2006 (2006), 062631. https://doi.org/10.1155/JAMSA/2006/62631 doi: 10.1155/JAMSA/2006/62631
    [20] R. B. Patel, Duality for multiobjective fractional variational control problems with (F, ρ)-convexity, Int. J. Stat. Manag. Syst., 3 (2000), 113–134. https://doi.org/10.1080/09720510.2000.10701010 doi: 10.1080/09720510.2000.10701010
    [21] T. Saeed, S. Treanţă, On sufficiency conditions for some robust variational control problems, Axioms, 12 (2023), 705. https://doi.org/10.3390/axioms12070705 doi: 10.3390/axioms12070705
    [22] T. Saeed, Robust optimality conditions for a class of fractional optimization problems, Axioms, 12 (2023), 673. https://doi.org/10.3390/axioms12070673 doi: 10.3390/axioms12070673
    [23] X. Sun, X. Feng, K. L. Teo, Robust optimality, duality and saddle points for multiobjective fractional semi-infinite optimization with uncertain data, Optim. Lett., 16 (2022), 1457–1476. https://doi.org/10.1007/s11590-021-01785-2 doi: 10.1007/s11590-021-01785-2
    [24] X. Sun, K. L. Teo, X. J. Long, Some characterizations of approximate solutions for robust semi-infinite optimization problems, J. Optim. Theory Appl., 191 (2021), 281–310. https://doi.org/10.1007/s10957-021-01938-4 doi: 10.1007/s10957-021-01938-4
    [25] X. Sun, W. Tan, K. L. Teo, Characterizing a class of robust vector polynomial optimization via sum of squares conditions, J. Optim. Theory Appl., 197 (2023), 737–764. https://doi.org/10.1007/s10957-023-02184-6 doi: 10.1007/s10957-023-02184-6
    [26] X. Sun, K. L. Teo, J. Zeng, X. L. Guo, On approximate solutions and saddle point theorems for robust convex optimization, Optim. Lett., 14 (2020), 1711–1730. https://doi.org/10.1007/s11590-019-01464-3 doi: 10.1007/s11590-019-01464-3
    [27] S. Treanţă, Efficiency in uncertain variational control problems, Neural. Comput. Appl., 33 (2021), 5719–5732. https://doi.org/10.1007/s00521-020-05353-0 doi: 10.1007/s00521-020-05353-0
    [28] S. Treanţă, Necessary and sufficient optimality conditions for some robust variational problems, Optim. Control Appl. Methods, 44 (2023), 81–90. https://doi.org/10.1002/oca.2931 doi: 10.1002/oca.2931
    [29] H. C. Wu, Duality theory for optimization problems with interval-valued objective functions, J. Optim. Theory Appl., 144 (2010), 615–628. https://doi.org/10.1007/s10957-009-9613-5 doi: 10.1007/s10957-009-9613-5
    [30] J. Zhang, S. Liu, L. Li, Q. Feng, The KKT optimality conditions in a class of generalized convex optimization problems with an interval-valued objective function, Optim. Lett., 8 (2014), 607–631. https://doi.org/10.1007/s11590-012-0601-6 doi: 10.1007/s11590-012-0601-6
  • This article has been cited by:

    1. Muhammad Uzair Awan, Sadia Talib, Muhammad Aslam Noor, Yu-Ming Chu, Khalida Inayat Noor, On post quantum estimates of upper bounds involving twice (p,q)-differentiable preinvex function, 2020, 2020, 1029-242X, 10.1186/s13660-020-02496-5
    2. Shu-Bo Chen, Saima Rashid, Muhammad Aslam Noor, Zakia Hammouch, Yu-Ming Chu, New fractional approaches for n-polynomial P-convexity with applications in special function theory, 2020, 2020, 1687-1847, 10.1186/s13662-020-03000-5
    3. Yu-Ming Chu, Muhammad Uzair Awan, Sadia Talib, Muhammad Aslam Noor, Khalida Inayat Noor, New post quantum analogues of Ostrowski-type inequalities using new definitions of left–right (p,q)-derivatives and definite integrals, 2020, 2020, 1687-1847, 10.1186/s13662-020-03094-x
    4. Saad Ihsan Butt, Muhammad Umar, Saima Rashid, Ahmet Ocak Akdemir, Yu-Ming Chu, New Hermite–Jensen–Mercer-type inequalities via k-fractional integrals, 2020, 2020, 1687-1847, 10.1186/s13662-020-03093-y
    5. Hua Wang, Humaira Kalsoom, Hüseyin Budak, Muhammad Idrees, Ahmet Ocak Akdemir, q -Hermite–Hadamard Inequalities for Generalized Exponentially s , m ; η -Preinvex Functions, 2021, 2021, 2314-4785, 1, 10.1155/2021/5577340
    6. Humaira Kalsoom, Muhammad Aamir Ali, Muhammad Idrees, Praveen Agarwal, Muhammad Arif, Adrian Neagu, New Post Quantum Analogues of Hermite–Hadamard Type Inequalities for Interval-Valued Convex Functions, 2021, 2021, 1563-5147, 1, 10.1155/2021/5529650
    7. Humaira Kalsoom, Miguel Vivas-Cortez, Muhammad Idrees, Praveen Agarwal, New Parameterized Inequalities for η-Quasiconvex Functions via (p, q)-Calculus, 2021, 23, 1099-4300, 1523, 10.3390/e23111523
    8. Saad Ihsan Butt, Hüseyin Budak, Muhammad Tariq, Muhammad Nadeem, Integral inequalities for n ‐polynomial s ‐type preinvex functions with applications , 2021, 44, 0170-4214, 11006, 10.1002/mma.7465
    9. Humaira Kalsoom, Miguel Vivas-Cortez, Muhammad Amer Latif, Trapezoidal-Type Inequalities for Strongly Convex and Quasi-Convex Functions via Post-Quantum Calculus, 2021, 23, 1099-4300, 1238, 10.3390/e23101238
    10. YONGFANG QI, GUOPING LI, FRACTIONAL OSTROWSKI TYPE INEQUALITIES FOR (s,m)-CONVEX FUNCTION WITH APPLICATIONS, 2023, 31, 0218-348X, 10.1142/S0218348X23501281
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(842) PDF downloads(34) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog