1. Introduction
Mathematical inequalities are fundamental in both pure and applied mathematics, offering essential tools in areas ranging from analysis to physics. Among these, the Hermite-Hadamard inequality for convex functions plays a pivotal role, which is defined as follows:
The Hermite-Hadamard inequality and a variety of refinements of Hermite-Hadamard inequality have been extensively studied by many researchers, including those involving Jensen and Mercer-type refinements.
Jensen inequality has been caught attention of many researshers, and many articles related to different versions of this inequality have been found in the literature. Jensen’s inequality can be given as follows:
Let
and
be non-negative weights such that
=1 if f is convex function on the interval [a, b], then
where
every and all
A new variant of Jensen inequality thast has been established by Mercer can be presented as follows:
In 2003, Mercer [1] proved another version of Jensen inequality, which is called Jensen-Mercer inequality and stated as follows.
Theorem 1.1
For a convex mapping
for following inequality holds for each
where
and
In 2013, Kian, et al. [2] used this new Jensen inequality and established the following new versions of Hermite-Hadamard inequality:
Theorem 1.2
For a convex mapping
for following inequality holds for each
and
and
The ordinary calculus of Newton and Leibniz is well known to be investigated extensively and intensively to produce a large number of related formulas and properties as well as applications in a variety of fields ranging from natural sciences to social sciences.
Recent studies have explored these inequalities using quantum and post-quantum calculus, which extend traditional calculus by means of q and (p,q)-analogues.
Quantum calculus, which is often known as q-calculus or calculus without limits, is based on finite difference. In quantum calculus we obtain q-analogues of mathematical objects which can be recaptured by taking. The history of q-calculus can be track to Euler, who first introduced q-calculus in the track of Newton’s work on infinite series.
Then, in 1910, F. H. Jackson presented a systematic study of q-calculus and defined the q-defined integral, which is known as the q-Jackson integral. In recent years, the interest in q-calculus been arising due to high demand of mathematics in this field. The q-calculus numerous applications in various fields of mathematics and other areas such as combinatory, dynamical systems, fractals, number theory, orthogonal polynomials, special functions, mechanics and also for scientific problems in some applied areas.
In 2013, Tariboon and Ntouyas defined new q-derivatives and q-integrals of a continuous function on a finite interval. These definitions have been studied in various inequalities, for example, Hermiter-Hadamard inequalities, Ostrowski inequalities, Fejér inequalities, Simpson inequalities and Newton inequalities, and the references cited therein [3-12].
Along with the development of the theory and application of q-calculus, the theory of q-calculus based on two parameters (p-q)-integral has also presented and received more attention during the last few decades.
Recent, Ali, et al. [3] and Sitthiwirattham, et al. [13] ued new techniques to prove the following two different and new versions of Hermite-Hadamard type inequalitites:
Theorem 1.3
For a convex mapping
for following inequality holds:
Remark
By setting the limit as q 1- in above Theorem, we recapture the traditional Hermite-Hadamard inequality.
Post-quantum calculus, also called (p,q)-calculus, is another generalization of q-calculus on the interval . The (p,q)-calculus consists of two-parameter quantum calculus (p and q-numbers) which are independent. The (p,q)- calculus was first introduced by Chakrabarti and Jagannathan in 1991. Then, the new (p,q)-deravative and (p,q)-integral of a continuous function on finite interval were by Tunc and Gov in 2016. In (p,q)-calculus, we obtain q-calculus formula for case p=1, and then get classical formula for case of . Base on (p,q)-calculus, many literatures have been published by many researchers, see [14-25] for more details and the references cited therein.
In this paper, we continue in this direction by developing Hermite-Hadamard-Mercer type inequalities in the framework of post-quantum calculus on coordinates. Our main contributions include novel identities for functions of two variables involving mixed partial (p1; p2; q1; q2) - derivatives and integrals.
2. Notation and preliminaries
The following is the brief introduction of the research of post-quantum calculus. Throughout this topic, we let p1; p2; q1; q2 be constants with
and
with
Then, for any real number, the (p1, q1) - analogue and (p2, q2) - analogue of m,n is defined by
and
which is generalization of the q_1-analogue such that
and
Definition 2.1 [26]
If is a continuous function, then (p, q) - derivative of the function on [a, b] by
with
Definition 2.2 [27]
If
is a continuous function, then (p, q)a - derivative of the function at x is defined by
with
For x = a, we state
if it exists and it is finite.
Definiiton 2.3 [16]
If
is a continuous function, then (p, q)b - derivative of the function at x is defined by
with
For x = b, we state
if it exists and it is finite.
Definition 2.4 [26]
If
is a continuous function and
then the (p, q) - integral is defined by
with
Definition 2.5 [27]
If
is a continuous function and
then the (p, q)a - integral is defined by
with
Definition 2.6 [16]
If
is a continuous function and
then the (p, q)b - integral is defined by
with
In [14], Ali et al. established the Hermite-Hadamard type inequlities on post quantum calculus.
Theorem 2.7
If
is a convex differentiable function on [a,b], then the (p, q)b - Hermite-Hadamard inequalities are as follows:
In [27], Tunc and Gov extend the Holdër inequalies ion post-quantum calculus.
Theorem 2.8
If
is a continuous function and r, s > 0 with
then
In [28], H. Kalsoom, et al. introduced the following notions of post-quantum partial derivatives:
Definition 2.9
Suppose that
is a continuous function of two variables. Then the derivatives are given by
for
for
for
and
for
Definition 2.10 [28]
Suppose that
is a continuous function of two variables. Then the definite (p1, p2, q1, q2) - integral are given by
for
Suppose that
is a continuous function of two variables. Then the definite (p1, p2, q1, q2) - integral are given by
for
Suppose that
is a continuous function of two variables. Then the definite (p1, p2, q1, q2) - integral are given by
for
and
Suppose that
is a continuous function of two variables. Then the definite (p1, p2, q1, q2) -integral are given by
for
Lemma 2.11 [25] (p1, p2, q1, q2) - Hölder’s inequality for functions of two variables
Let f, g be (p1, p2, q1, q2) - integrable functions on
and
with r, s > 1. Then, we have
for all
Lemma 2.12 [25] (p1, p2, q1, q2) - power mean inequality for functions of two variables
Let f, g be (p1, p2, q1, q2) - integrable functions on
and 1. Then, we have
for all
3. Main results
Lemma 3.1
Let
be a twice partially (p1, p2, q1, q2) - differentiable function on
If
are continuous and (p1, p2, q1, q2) -integrable on
Then we have
where
for
and
Proof:
By the definition 2.9, we have
Moreover,
By definition 2.10, we have
So that,
Similarly, by the equality, we obtain the identities
Thus, we have
Corollary 3.2
If we set p1, p2 = 1, then the Lemma 3.1 reduces to the following equality.
where
Theorem 3.3
If the conditions of Lemma 3.1 hold and
are coordinated convex, then we have the following inequality:
proof:
From the result of lemma 3.1 and Jensen-Mercer inequality, we have
Corollary 3.4
If we set p1, p2 = 1, then the Theorem 3.2 reduces to the following inequality.
Theorem 3.5
Suppose that tha assumptions of Lemma 3.1 are hold. If
are coordinated convex function on
, then we have the inequality
where
Proof:
From the Lemma 3.1 and and Jensen-Mercer inequality by using the Hölder inequality and the convexity of
we obtain
Theorem 3.6
Suppose that tha assumptions of Lemma 3.1 are hold. If
are coordinated convex function on
for
then we have the inequality
Proof:
From the Lemma 3.1 and and Jensen-Mercer inequality by using the power-mean inequality and the convexity of
we obtain
Conclusion
In this paper, we have established several new Hermite-Hadamard-Mercer type inequalities on coordinates using post-quantum (p, q)-calculus. These results generalize existing inequalities and provide a unified framework encompassing both classical and quantum cases. Our work extends the literature by presenting double-integral identities under (p1, p2, q1, q2) - partial differentiability and integrability assumptions.Further research may explore: Higher-dimensional generalizations, Inequalities for other convexity types (e.g., s-convex, log-convex) and the applications in optimization or quantum information theory.
Data availability
All data required for this paper is included within this paper.