Next Article in Journal
On the Potential Vector Fields of Soliton-Type Equations
Next Article in Special Issue
Qualitative Outcomes on Monotone Iterative Technique and Quasilinearization Method on Time Scale
Previous Article in Journal
Stability of Fixed Points of Partial Contractivities and Fractal Surfaces
Previous Article in Special Issue
Fractional-Order Sequential Linear Differential Equations with Nabla Derivatives on Time Scales
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

On Some New Dynamic Hilbert-Type Inequalities across Time Scales

1
Department of Mathematics, College of Science, King Khalid University, P.O. Box 9004, Abha 61413, Saudi Arabia
2
Mathematical Institute, Slovak Academy of Sciences, Grekošákova 6, 04001 Košice, Slovakia
3
Department of Mathematics, Faculty of Science, Benha University, Benha 13511, Egypt
4
Department of Mathematics, College of Arts and Sciences, King Khalid University, P.O. Box 64512, Abha 62529, Sarat Ubaidah, Saudi Arabia
5
Department of Mathematics, Faculty of Science, Al-Azhar University, Nasr City 11884, Egypt
*
Author to whom correspondence should be addressed.
Axioms 2024, 13(7), 475; https://doi.org/10.3390/axioms13070475
Submission received: 19 May 2024 / Revised: 29 June 2024 / Accepted: 12 July 2024 / Published: 14 July 2024
(This article belongs to the Special Issue Infinite Dynamical System and Differential Equations)

Abstract

:
In this article, we present some novel dynamic Hilbert-type inequalities within the framework of time scales T . We achieve this by utilizing Hölder’s inequality, the chain rule, and the mean inequality. As specific instances of our findings (when T = N and T = R ), we obtain the discrete and continuous analogues of previously established inequalities. Additionally, we derive other inequalities for different time scales, such as T = q N 0 for q > 1 , which, to the best of the authors’ knowledge, is a largely novel conclusion.

1. Introduction

In the early 1900s, the renowned mathematician David Hilbert formulated his famous inequality, known as the double series Hilbert inequality (see [1]), wherein he established that if { G m } m = 1 ,   { Z n } n = 1 inf are two real sequences, such that 0 < m = 1 G m 2 < and 0 < n = 1 Z n 2 < , then
n = 1 m = 1 G m Z n m + n 2 π m = 1 G m 2 1 2 n = 1 Z n 2 1 2 .
In 1911, Schur demonstrated in his paper [2] that the constant π in (1) is optimal. Furthermore, he established the integral analogue of (1), which is now recognized as the Hilbert integral inequality in the form
0 0 S ( x ) T ( y ) x + y d x d y π 0 S 2 ( x ) d x 1 2 0 T 2 ( y ) d y 1 2 ,
where S and T are real functions such that 0 < 0 inf S 2 ( x ) d x < ,   0 < 0 T 2 ( y ) d y < , and π in (2) is still the best possible constant factor.
The inequalities expressed as (1) and (2) are crucial in the theory and application of integral inequalities, especially in analyzing both the qualitative and quantitative aspects of solutions to differential and integral equations. Recently, there has been rapid development in fractal theory, which has found widespread use in science and engineering. Some researchers have utilized fractal theory and weight function methods to generalize classical inequalities effectively. For instance, Liu [3] established a Hilbert-type integral inequality and its equivalent form on a fractal set. Hilbert-type inequalities play a significant role in mathematics, particularly in complex and numerical analysis. Over the years, these inequalities have seen numerous refinements, generalizations, extensions, and applications in the literature (see [4,5,6,7]).
In 1925, Hardy [8] extended (1) by introducing a pair of conjugate exponents ( η , λ ) , where η , λ > 1 and satisfying 1 / η + 1 / λ = 1 , as follows. If G m ,   Z n 0 , such that 0 < m = 1 G m η < , and 0 < n = 1 Z n λ < , then
n = 1 m = 1 G m Z n m + n π sin π η m = 1 G m η 1 η n = 1 Z n λ 1 λ .
In [9], the authors established the equivalent integral form of (3) as
0 0 S ( x ) T ( y ) x + y d x d y π sin π η 0 S η ( x ) d x 1 η 0 T λ ( y ) d y 1 λ ,
where S ,   T 0 , such that 0 < 0 S p ( x ) d x < and 0 < 0 T λ ( y ) d y < . The constant factor π / sin ( π / η ) in (3) and (4) is optimal.
In 1998, Pachpatte [10] presented a new inequality akin to the Hilbert inequality as follows: let G ( s ) : 0 , 1 , 2 , , p N R and Z ( ϑ ) : 0 , 1 , 2 , , q N R with G ( 0 ) = Z ( 0 ) = 0 . Define the operators as G s = G s G s 1 ,   Z ϑ = Z ϑ Z ϑ 1 . Then
s = 1 p ϑ = 1 q G s Z ϑ s + ϑ 1 2 p q s = 1 p ( p s + 1 ) G s 2 1 2 × ϑ = 1 q ( q ϑ + 1 ) Z ϑ 2 1 2 .
In 2000, Pachpatte [11] generalized (5) by introducing one pair of conjugate exponents ( η , μ ) : η , μ > 1 with 1 / η + 1 / μ = 1 and proved that if G ( s ) : 0 , 1 , 2 , , p N R and Z ( ϑ ) : 0 , 1 , 2 , , q N R with G ( 0 ) = Z ( 0 ) = 0 , then
s = 1 p ϑ = 1 q G s Z ϑ μ s η 1 + η ϑ μ 1 1 η μ p η 1 η q μ 1 μ s = 1 p ( p s + 1 ) G s η 1 η × ϑ = 1 q ( q ϑ + 1 ) Z ϑ μ 1 μ .
In 2002, Kim et al. [12] extended (6) and demonstrated that if η , μ > 1 ,   G ( s ) : 0 , 1 , 2 , , p N R and Z ( ϑ ) : 0 , 1 , 2 , , q N R with G ( 0 ) = Z ( 0 ) = 0 , then
s = 1 p ϑ = 1 q G s Z ϑ μ s η 1 η + μ η μ + η ϑ μ 1 η + μ η μ 1 η + μ p η 1 η q μ 1 μ s = 1 p ( p s + 1 ) G s η 1 η × ϑ = 1 q ( q ϑ + 1 ) Z ϑ μ 1 μ .
Also, the authors [12] established the continuous analogue of (7) as follows: if η , μ > 1 and S ( θ ) , and T ( ϑ ) are real continuous functions on 0 , x ,   0 , y , respectively, with S ( 0 ) = T ( 0 ) = 0 , then for x , y 0 , , we have
0 x 0 y S ( θ ) T ( ϑ ) μ θ η 1 η + μ η μ + η ϑ μ 1 η + μ η μ d θ d ϑ 1 η + μ x η 1 η y μ 1 μ 0 x x θ S ( θ ) η d θ 1 η 0 y y ϑ T ( ϑ ) μ d ϑ 1 μ .
In 2011, Chang-Jian et al. [13] generalized (5) and demonstrated that if λ i > 1 , such that 1 / λ i + 1 / q i = 1 , G i ( s i ) is a real sequence defined for s i = 0 , 1 , 2 , , m i , where m i is a natural number and G i 0 = 0 ,   i = 1 , 2 , , n . . Define the operator ∇ by G i ( s i ) = G i ( s i ) G i ( s i 1 ) for any function G i ( s i ) ,   i = 1 , 2 , , n . Then
s 1 = 1 m 1 s 2 = 1 m 2 s n = 1 m n i = 1 n G i ( s i ) i = 1 n s i / q i i = 1 n 1 / q i M i = 1 n s i = 1 m i m i s i + 1 G i ( s i ) λ i 1 λ i ,
where
M = n i = 1 n 1 λ i i = 1 n 1 / λ i n . i = 1 n m i 1 / q i .
Also, the authors of [13] proved that if h i 1 and λ i > 1 are constants such that 1 / λ i + 1 / q i = 1 , T i ( s i ) is a real valued differentiable function defined on [ 0 , x i ) , where x i 0 , . Assume T i 0 = 0 for i = 1 , 2 , , n . Then
0 x 1 0 x n i = 1 n T i h i ( s i ) i = 1 n s i / q i i = 1 n 1 / q i d s n d s 1 K i = 1 n 0 x i x i s i T i h i 1 ( s i ) . T i ( s i ) λ i d s i 1 λ i ,
where
K = n i = 1 n 1 λ i i = 1 n 1 / λ i n . i = 1 n h i x i 1 / q i .
Also, they demonstrated that if λ i , q i > 1 , such that 1 / λ i + 1 / q i = 1 , G i ( s i , t i ) is a real sequence defined for s i , t i , where s i = 0 , 1 , 2 , , m i ,   t i = 0 , 1 , 2 , , n i and m i , n i   i = 1 , 2 , , n are natural numbers and assuming that G i ( 0 , t i ) = G i ( s i , 0 ) = 0 for all i = 1 , 2 , , n . Define the operator 1 and 2 as
H 1 G i ( s i , t i ) = G i ( s i , t i ) G i ( s i 1 , t i ) ,
2 G i ( s i , t i ) = G i ( s i , t i ) G i ( s i , t i 1 ) .
Then
s 1 = 1 m 1 t 1 = 1 n 1 s n = 1 m n t n = 1 n n i = 1 n G i ( s i , t i ) i = 1 n s i t i / q i i = 1 n 1 / q i L i = 1 n s i = 1 m i t i = 1 n i m i s i + 1 n i t i + 1 2 1 G i ( s i , t i ) λ i 1 λ i ,
where
L = n i = 1 n 1 λ i i = 1 n 1 / λ i n . i = 1 n m i n i 1 / q i .
In the last few decades, much attention has been devoted to establishing discrete analogues of the corresponding continuous results in various fields of analysis. This appears along with establishing a dynamic inequality in this paper by using a general domain called a time scale T . A time scale T is an arbitrary non-empty closed subset of the real numbers R . For more details about dynamic inequalities and applications on time scales, see [14,15,16,17,18,19].
The aim of this paper is to prove similar analogues of the inequalities (8), (9) on time scales, and we can also generalize (10) on time scale delta calculus for an increasing function by establishing some new dynamic Hilbert-type inequalities on time scale delta calculus.
The remainder of this paper is organized as follows. In Section 2, we show some basics of the time scale theory and some lemmas on time scales needed in Section 3, where we prove our results. These results as special cases when T = N and T = R give the inequalities ((8) and (10)), (9), respectively. Also, we can obtain other inequalities on different time scales, like T = q N for q > 1 .

2. Preliminaries and Basic Lemmas

In 2001, Bohner and Peterson [20] defined the forward jump operator by σ ( τ ) : = inf { s T : s > τ } . For any function S : T R , the notation S σ ( τ ) denotes S ( σ ( τ ) ) . We define the time scale interval [ a , b ] T by [ a , b ] T : = [ a , b ] T .
In the following, we state the definition of rd−continuous and Δ−derivative function.
Definition 1
([20]). A function S : T R is called rd−continuous provided it is continuous at right-dense points in T and its left-sided limits exist (finite) at left-dense points in T . The set of rd−continuous functions S : T R is denoted by C r d ( T , R ) .
Definition 2
([20]). Assume that S : T R is a function and let t T . We define S Δ ( t ) to be the number, provided it exists, as follows: for any ε > 0 , there is a neighborhood U of t ,   U = ( t δ ,   t + δ ) T for some δ > 0 , such that
| S ( σ ( t ) ) S ( γ ) S Δ ( t ) ( σ ( t ) γ ) | ε | σ ( t ) γ | for all γ U , γ σ ( t ) .
In this case, we say S Δ ( t ) is the delta or Hilger derivative of S at t.
In the following, we state several values of Δ —differentiable function at a point t T .
Theorem 1
([20]). Assume S : T R is a function and let t T k . Then we have the following.
1.
If S is differentiable at t, then f is continuous at t.
2.
If S is continuous at t and t is right-scattered, then S is differentiable at t with
S Δ ( t ) = S ( σ ( t ) ) S ( t ) σ ( t ) t .
3.
If t is right-dense, then S is differentiable if the limit
lim γ t S ( t ) S ( γ ) t γ ,
exists as a finite number. In this case,
S Δ ( t ) = lim γ t S ( t ) S ( γ ) t γ .
Example 1.
1. If T = R , then for S : R R , we obtain
S Δ ( t ) = lim γ t S ( t ) S ( γ ) t γ = S ( t ) ,
where S is the usual derivative.
2. If T = N , then σ ( t ) = t + 1 , and for S : N R , we have
S Δ ( t ) = S ( σ ( t ) ) S ( t ) σ ( t ) t = S ( t + 1 ) S ( t ) 1 = Δ S ( t ) ,
where Δ is the usual forward difference operator.
3. If T = { t : t = q k ,   k N 0 ,   q > 1 } , then we have σ ( t ) = q t and
S Δ ( t ) = Δ q S ( t ) = S ( q t ) S ( t ) ( q 1 ) t .
The following theorem is about the chain rule formula on time scales.
Theorem 2
(Chain Rule [20] Theorem 1.87). Assume T : R R is continuous, T : T R is delta-differentiable on T and S : R R is continuously differentiable. Then γ exists in the real interval [ t , σ ( t ) ] with
S T Δ ( t ) = S T ( γ ) T Δ ( t ) .
Definition 3
([20]). A function S : T R is called an antiderivative of s : T R , provided that
S Δ ( t ) = s ( t ) holds for all t T k .
In this case, the Cauchy integral of s is defined by
r α s ( t ) Δ t = S ( α ) S ( r ) , for all r , α T .
Theorem 3
([20]). Every rd–continuous function S : T R has an antiderivative. In particular, if t 0 T , then
t 0 t S ( τ ) Δ τ Δ = S ( t ) , for t T .
In the following, we present the properties of integration on time scales.
Theorem 4
([20]). If a ,   b ,   c T ,   α ,   β R and S ,   T C r d ( [ a ,   b ] T , R ) , then
1.
a b α S ( t ) + β T ( t ) Δ t = α a b S ( t ) Δ t + β a b T ( t ) Δ t .
2.
a b S ( t ) Δ t = b a S ( t ) Δ t .
3.
a b S ( t ) Δ t = a c S ( t ) Δ t + c b S ( t ) Δ t .
4.
a a S ( t ) Δ t = 0 .
5.
a b S ( t ) Δ t a b S ( t ) Δ t .
6.
If S ( t ) 0 for all t [ a , b ] T , then a b S ( t ) Δ t 0 .
Theorem 5
([21]). Let a ,   b T and S C r d ( T , R ) . Then, the following properties hold:
( i ) If T = R , then
a b S ( t ) Δ t = a b S ( t ) d t .
( i i ) If T = N 0 { 0 } , then
a b S ( t ) Δ t = t = a b 1 S ( t ) .
( i i i ) If T = { t : t = q k , k N 0 ,   q > 1 } , then
t 0 S ( t ) Δ t = k = n 0 S ( q k ) μ ( q k ) .
In the following, we present some auxiliary lemmas that we need to prove our results.
Lemma 1
(Integration by Parts [21]). If a , b T and S ,   T C r d ( [ a , b ] T ,   R ) , then
a b S ( t ) T Δ ( t ) Δ t = S ( t ) T ( t ) a b a b S Δ ( t ) T σ ( t ) Δ t .
Lemma 2
(lHölder’s Inequality [21,22]). If a ,   b T and S ,   T C r d ( [ a , b ] T ,   R ) , then
a b | S ( t ) T ( t ) | Δ t a b | S ( t ) | γ Δ t 1 γ a b | T ( t ) | ν Δ t 1 ν ,
where γ > 1 and 1 / γ + 1 / ν = 1 . The inequality (13) is reversed for 0 < γ < 1 or γ < 0 .
Let T 1 and T 2 be time scales. Assume that C C r d denotes the set of functions S t 1 , t 2 on T 1 × T 2 , where S is r d continuous in t 1 and t 2 . Let C C r d denote the set of all functions C C r d for which both the Δ 1 partial derivative with respect to t 1 and the Δ 2 partial derivative with respect to t 2 exist and are in C C r d .
Lemma 3
(Two dimensional Hölder’s inequality [23] Theorem 3.3). Assume that a ,   b T with a < b ,   S ,   T C C r d ( a , b T × a , b T , R ) and γ ,   ν > 1 such that 1 / γ + 1 / ν = 1 . Then
a b a b | S ( τ , ξ ) T ( τ , ξ ) | Δ 1 τ Δ 2 ξ a b a b | S ( τ , ξ ) | γ Δ 1 τ Δ 2 ξ 1 γ a b a b | T ( τ , ξ ) | ν Δ 1 τ Δ 2 ξ 1 ν .
Lemma 4
(Fubini’s theorem [24]). If a , b , c , d T and S C C r d a , b T × c , d T , R is Δ−integrable, then
a b c d S x , y Δ 2 y Δ 1 x = c d a b S x , y Δ 1 x Δ 2 y .
Lemma 5
(Mean inequality [9]). If α i , β i > 0 for i = 1 , 2 , , n , then
i = 1 n α i β i i = 1 n α i β i i = 1 n β i i = 1 n β i i = 1 n β i .
Lemma 6.
Let p i , r i > 1 with 1 / p i + 1 / r i = 1 and s i > 0 , where i = 1 , 2 , , n . Then
i = 1 n s i 1 / r i i = 1 n s i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i .
Proof. 
Applying Lemma 5 with α i = s i and β i = 1 / r i , we observe that
i = 1 n s i 1 / r i i = 1 n s i / r i i = 1 n 1 / r i i = 1 n 1 / r i i = 1 n 1 / r i .
Since 1 / r i = 1 1 / p i , we can obtain that
i = 1 n 1 / r i = i = 1 n 1 1 / p i = n i = 1 n 1 / p i ,
and then the inequality (17) becomes
i = 1 n s i 1 / r i i = 1 n s i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i ,
which is (16). □

3. Main Results

In this section, we present the key results of our study. Firstly, we establish the time scale version of (8).
Theorem 6.
Let a i , ε i T ,   p i , r i > 1 , such that 1 / p i + 1 / r i = 1 and let λ i C r d ( a i , ε i T ,   R ) be a delta-differentiable function with λ i ( a i ) = 0 ;   i = 1 , 2 , , n . Then
a n ε n a 1 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n M i = 1 n a i ε i ε i σ ( ξ i ) λ i Δ ( ξ i ) p i Δ ξ i 1 p i ,
where
M = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i a i 1 q i .
Proof. 
Applying the property (5) of Theorem 4 and the hypothesis λ i ( a i ) = 0 , we obtain
a i ξ i λ i Δ ( t i ) Δ t i a i ξ i λ i Δ ( t i ) Δ t i = λ i ( ξ i ) λ i ( a i ) = λ i ( ξ i ) ,
and then
i = 1 n λ i ( ξ i ) i = 1 n a i ξ i λ i Δ ( t i ) Δ t i .
Applying (13) on a i ξ i λ i Δ ( t i ) Δ t i with f ( t i ) = λ i Δ ( t i ) and g ( t i ) = 1 , we observe that
a i ξ i λ i Δ ( t i ) Δ t i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i a i ξ i Δ t i 1 r i = ξ i a i 1 r i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i ,
and then
i = 1 n a i ξ i λ i Δ ( t i ) Δ t i i = 1 n ξ i a i 1 r i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i = i = 1 n ξ i a i 1 r i i = 1 n a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i .
Substituting (21) into (20), we obtain
i = 1 n λ i ( ξ i ) i = 1 n ξ i a i 1 r i i = 1 n a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i .
Applying (16) with s i = ξ i a i , we have
i = 1 n ξ i a i 1 / r i i = 1 n ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i .
Substituting (23) into (22), we obtain
i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i i = 1 n a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i .
Dividing (24) by i = 1 n ξ i a i / r i i = 1 n 1 / r i and integrating over ξ i from a i to ε i ,   i = 1 , 2 , , n , we observe that
a n ε n a 1 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n a n ε n a 1 ε 1 i = 1 n a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ 1 Δ ξ n = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i .
Applying (13) on a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i with f ξ i = a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i and g ξ i = 1 , we have
a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i a i ε i Δ ξ i 1 r i = ε i a i 1 r i a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i ,
and then
i = 1 n a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i i = 1 n ε i a i 1 r i a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i = i = 1 n ε i a i 1 r i i = 1 n a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i .
Substituting (26) into (25), we have
a n ε n a 1 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i a i 1 r i i = 1 n a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i .
Applying (12) on a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i with f ( ξ i ) = a i ξ i λ i Δ ( t i ) p i Δ t i and g Δ ( ξ i ) = 1 , we obtain
a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i = a i ξ i λ i Δ ( t i ) p i Δ t i g ( ξ i ) a i ε i a i ε i λ i Δ ( ξ i ) p i g σ ( ξ i ) Δ ξ i ,
where g ( ξ i ) = ξ i ε i . Since g ( ε i ) = 0 , we can find from (28) that
a i ε i a i ξ i λ i Δ ( t i ) p i Δ t i Δ ξ i = a i ε i λ i Δ ( ξ i ) p i ε i σ ( ξ i ) Δ ξ i .
Substituting (29) into (27), we obtain
a n ε n a 1 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i a i 1 r i × i = 1 n a i ε i ε i σ ( ξ i ) λ i Δ ( ξ i ) p i Δ ξ i 1 p i .
Substituting (19) into (30), we obtain
a n ε n a 1 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n M i = 1 n a i ε i ε i σ ( ξ i ) λ i Δ ( ξ i ) p i Δ ξ i 1 p i ,
which is (18). □
Corollary 1.
If we put T = N 0 and a i = 0 for i = 1 , 2 , , n , into Theorem 6, then σ ( ξ i ) = ξ i + 1 , and we obtain the analogue of inequality (8) as follows
ξ 1 = 0 ε 1 1 ξ n = 0 ε n 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i / r i i = 1 n 1 / r i M i = 1 n ξ i = 0 ε i 1 ε i ξ i 1 Δ λ i ( ξ i ) p i 1 p i ,
where
M = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i 1 r i .
In the following, we present some special cases in (the continuous and quantum) calculie, i.e., when T = R and T = q N 0 for q > 1 . These cases are new and interesting for the reader.
Corollary 2.
In Theorem 6, if T = R ,   a i = 0 ,   p i , r i > 1 , such that 1 / p i + 1 / r i = 1 and λ i C ( 0 , ε i ,   R ) is a differentiable function with λ i ( 0 ) = 0 ;   i = 1 , 2 , , n , then σ ( ξ i ) = ξ i and we obtain
0 ε n 0 ε 1 i = 1 n λ i ( ξ i ) i = 1 n ξ i / r i i = 1 n 1 / r i d ξ 1 d ξ n M i = 1 n 0 ε i ε i ξ i λ i ( ξ i ) p i d ξ i 1 p i ,
where
M = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i 1 q i .
Corollary 3.
In Theorem 6, if T = q N 0 for q > 1 ,   p i , r i > 1 , such that 1 / p i + 1 / r i = 1 and λ i C ( a i , ε i T ,   R ) with λ i ( a i ) = 0 ;   i = 1 , 2 , , n , then σ ( ξ i ) = q ξ i and we obtain
ξ 1 = a 1 ε 1 / q ξ n = a n ε n / q i = 1 n q 1 ξ i λ i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i M i = 1 n ξ i = a i ε i / q q 1 ξ i ε i q ξ i Δ q λ i ( ξ i ) p i 1 p i ,
where
Δ q λ i ( ξ i ) = λ i ( q ξ i ) λ i ( ξ i ) q 1 ξ i ,
and
M = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i a i 1 r i .
In the following theorem, we generalize the previous results for two variables.
Theorem 7.
Assume that a i , ε i , ϵ i T ,   p i , r i > 1 , such that 1 / p i + 1 / r i = 1 and λ i C C r d ( a i , ε i T × a i , ϵ i T ,   R ) with λ i ( a i , ξ i ) = λ i ( τ i , a i ) = 0 for ξ i a i , ε i T and τ i a i , ϵ i T ,   i = 1 , 2 , , n . Then
a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n N i = 1 n a i ε i a i ϵ i ϵ i σ τ i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 1 τ i Δ 2 ξ i 1 p i ,
where
N = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i a i 1 r i ε i a i 1 r i .
Here, the Δ 1 —derivative of the function λ ( τ , ξ ) is the Δ —derivative with respect to the first variable τ and the Δ 2 —derivative of the function λ ( τ , ξ ) is the Δ —derivative with respect to the second variable ξ .
Proof. 
Applying the property (5) of Theorem 4, Fubini’s theorem and the hypothesis λ i ( a i , ξ i ) = λ i ( τ i , a i ) = 0 , we obtain
a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i = a i ξ i a i τ i λ i Δ 2 ( t i , ϑ i ) Δ 1 Δ 1 t i Δ 2 ϑ i = λ i ( τ i , ξ i ) λ i ( τ i , a i ) λ i ( a i , ξ i ) + λ i ( a i , a i ) = λ i ( τ i , ξ i ) ,
and then
i = 1 n λ i ( τ i , ξ i ) i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i .
Applying (14) on a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i with h ( t i , ϑ i ) = 1 ,   f ( t i , ϑ i ) = 1 and g ( t i , ϑ i ) = λ i Δ 2 Δ 1 ( t i , ϑ i ) , we observe
a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i τ i a i 1 r i ξ i a i 1 r i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i ,
and then
i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) Δ 2 ϑ i Δ 1 t i i = 1 n τ i a i 1 r i ξ i a i 1 r i i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i .
Substituting (35) into (34), we observe that
i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i 1 r i ξ i a i 1 r i i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i .
Applying (16) with s i = τ i a i ξ i a i , we have
i = 1 n τ i a i 1 r i ξ i a i 1 r i i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i .
Substituting (37) into (36), we obtain
i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i × i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i .
Dividing (38) by i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i and integrating over ξ i and τ i from a i to ε i and ϵ i for i = 1 , 2 , , n , respectively, we observe that
a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n n i = 1 n 1 / p i i = 1 n 1 / p i n × a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n = n i = 1 n 1 / p i i = 1 n 1 / p i n × i = 1 n a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i Δ 2 ξ i Δ 1 τ i .
Applying (14) on a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i Δ 2 ξ i Δ 1 τ i with h ( ξ i , τ i ) = 1 ,   f ( ξ i , τ i ) = 1 and
g ( ξ i , τ i ) = a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i ,
we have
a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i Δ 2 ξ i Δ 1 τ i ϵ i a i 1 r i ε i a i 1 r i a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 2 ξ i Δ 1 τ i 1 p i ,
and then
i = 1 n a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i 1 p i Δ 2 ξ i Δ 1 τ i i = 1 n ϵ i a i 1 r i ε i a i 1 r i i = 1 n a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 2 ξ i Δ 1 τ i 1 p i .
Substituting (40) into (39) and applying the Fubini theorem, we obtain
a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i a i 1 r i ε i a i 1 r i × i = 1 n a i ϵ i a i ε i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 2 ξ i Δ 1 τ i 1 p i = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i a i 1 r i ε i a i 1 r i × i = 1 n a i ε i a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i Δ 2 ξ i 1 p i .
Applying (12) on a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i , with
f ( τ i ) = a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i and g Δ ( τ i ) = 1 ,
we obtain
a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i = g ( τ i ) a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i a i ϵ i a i ϵ i g σ ( τ i ) a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 1 τ i ,
where g ( τ i ) = τ i ϵ i . Since g ϵ i = 0 , we know from (42) that
a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i = a i ϵ i ϵ i σ τ i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 1 τ i .
Integrating (43) over ξ i from a i to ε i and then applying Fubini’s theorem, we obtain
a i ε i a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i Δ 2 ξ i = a i ε i a i ϵ i ϵ i σ τ i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 1 τ i Δ 2 ξ i = a i ϵ i a i ε i ϵ i σ τ i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 2 ξ i Δ 1 τ i = a i ϵ i ϵ i σ τ i a i ε i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 2 ξ i Δ 1 τ i .
Applying (12) on a i ε i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 2 ξ i with f ξ i = a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i and g Δ ξ i = 1 , we see
a i ε i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 2 ξ i = g ξ i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i a i ε i a i ε i g σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 2 ξ i ,
where g ξ i = ξ i ε i . Since g ε i = 0 , we know from (45) that
a i ε i a i ξ i λ i Δ 2 Δ 1 ( τ i , ϑ i ) p i Δ 2 ϑ i Δ 2 ξ i = a i ε i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 2 ξ i .
Substituting (46) into (44) and applying Fubini’s theorem, we obtain
a i ε i a i ϵ i a i τ i a i ξ i λ i Δ 2 Δ 1 ( t i , ϑ i ) p i Δ 2 ϑ i Δ 1 t i Δ 1 τ i Δ 2 ξ i = a i ϵ i ϵ i σ τ i a i ε i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 2 ξ i Δ 1 τ i = a i ε i a i ϵ i ϵ i σ τ i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 1 τ i Δ 2 ξ i .
Substituting (47) into (41), we see that
a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i a i 1 r i ε i a i 1 r i × i = 1 n a i ε i a i ϵ i ϵ i σ τ i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 1 τ i Δ 2 ξ i 1 p i .
Substituting (33) into (48), we have
a n ϵ n a 1 ϵ 1 a n ε n a 1 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i Δ 2 ξ 1 Δ 2 ξ n Δ 1 τ 1 Δ 1 τ n N i = 1 n a i ε i a i ϵ i ϵ i σ τ i ε i σ ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i Δ 1 τ i Δ 2 ξ i 1 p i ,
which is (32). □
Corollary 4.
If T = N 0 and a i = 0 , then σ ξ i = ξ i + 1 and we obtain the analogue of inequality (10) as follows:
τ 1 = 0 ϵ 1 1 τ n = 0 ϵ n 1 ξ 1 = 0 ε 1 1 ξ n = 0 ε n 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i ξ i / r i i = 1 n 1 / r i N i = 1 n ξ i = 0 ε i 1 τ i = 0 ϵ i 1 ϵ i τ i 1 ε i ξ i 1 Δ 2 Δ 1 λ i ( τ i , ξ i ) p i 1 p i ,
where Δ 1 λ i ( τ i , ξ i ) = λ i ( τ i + 1 , ξ i ) λ i ( τ i , ξ i ) ,   Δ 2 λ i ( τ i , ξ i ) = λ i ( τ i , ξ i + 1 ) λ i ( τ i , ξ i ) and
N = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ε i ϵ i 1 r i .
In the following corollaries, we show some particular cases in (the continuous and quantum) calculie, i.e., when T = R and T = q N 0 for q > 1 , which are original.
Corollary 5.
If T = R ,   a i = 0 ,   p i , r i > 1 , such that 1 / p i + 1 / r i = 1 ,   λ i C ( 0 , ε i × [ 0 , ϵ i ] ,   R ) with λ i ( 0 , ξ i ) = λ i ( τ i , 0 ) = 0 for ξ i 0 , ε i and τ i 0 , ϵ i ,   i = 1 , 2 , , n , then σ ξ i = ξ i and we obtain
0 ϵ n 0 ϵ 1 0 ε n 0 ε 1 i = 1 n λ i ( τ i , ξ i ) i = 1 n τ i ξ i / r i i = 1 n 1 / r i d ξ 1 d ξ n d τ 1 d τ n N i = 1 n 0 ε i 0 ϵ i ϵ i τ i ε i ξ i 2 ξ i τ i λ i ( τ i , ξ i ) p i d τ i d ξ i 1 p i ,
where
N = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i ε i 1 r i .
Corollary 6.
If T = q N 0 for q > 1 ,   p i , r i > 1 such that 1 / p i + 1 / r i = 1 and λ i : a i , ε i T × a i , ϵ i T R with λ i ( a i , ξ i ) = λ i ( τ i , a i ) = 0 for ξ i a i , ε i T and τ i a i , ϵ i T ,   i = 1 , 2 , , n , then σ t = q t and we obtain
τ 1 = a 1 ϵ 1 / q τ n = a n ϵ n / q ξ 1 = a 1 ε 1 / q ξ n = a n ε n / q i = 1 n q 1 2 τ i ξ i λ i ( τ i , ξ i ) i = 1 n τ i a i ξ i a i / r i i = 1 n 1 / r i N i = 1 n ξ i = a i ε i / q τ i = a i ϵ i / q ϵ i q τ i ε i q ξ i λ i Δ 2 Δ 1 ( τ i , ξ i ) p i 1 p i ,
where
N = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n ϵ i a i 1 q i ε i a i 1 q i .
Theorem 8.
Let a i , ε i T ,   h i 1 ,   p i , r i > 1 such that 1 / p i + 1 / r i = 1 and λ i C r d ( a i , ε i T ,   R + { 0 } ) is a delta-differentiable function and an increasing function with λ i ( a i ) = 0 ;   i = 1 , 2 , , n . Then
a 1 ε 1 a n ε n i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n Q i = 1 n a i ε i ε i σ ( ξ i ) λ i σ ( ξ i ) h i 1 λ i Δ ( ξ i ) p i Δ ξ i 1 p i ,
where
Q = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n h i ε i a i 1 r i .
Proof. 
Applying the chain rule formula (11) on the term λ i h i ( t i ) ,   h i 1 , we obtain
λ i h i ( t i ) Δ = h i λ i h i 1 ( ζ i ) λ i Δ ( t i ) ,
where ζ i t i , σ ( t i ) . Since λ i is an increasing function, h i 1 and ζ i σ ( t i ) , we know from (51) that
λ i h i ( t i ) Δ h i λ i σ ( t i ) h i 1 λ i Δ ( t i ) ,
and then (where λ i ( a i ) = 0 ), we observe that
h i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) Δ t i a i ξ i λ i h i ( t i ) Δ Δ t i = λ i h i ( ξ i ) λ i h i ( a i ) = λ i h i ( ξ i ) .
Thus,
i = 1 n h i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) Δ t i i = 1 n λ i h i ( ξ i ) .
Applying (13) on a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) Δ t i with f ( t i ) = λ i σ ( t i ) h i 1 λ i Δ ( t i ) and g ( t i ) = 1 , we have
a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) Δ t i a i ξ i Δ t i 1 r i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i = ξ i a i 1 r i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i ,
and then
i = 1 n a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) Δ t i i = 1 n ξ i a i 1 r i i = 1 n a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i .
Substituting (53) into (52), we get
i = 1 n λ i h i ( ξ i ) i = 1 n h i ξ i a i 1 r i i = 1 n a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i .
Applying (16) with s i = ξ i a i , we have
i = 1 n ξ i a i 1 / r i i = 1 n ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i .
Substituting (55) into (54), we obtain
i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i n i = 1 n 1 / p i n i = 1 n 1 / p i × i = 1 n h i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i .
Dividing (56) by i = 1 n ξ i a i / r i i = 1 n 1 / r i and integrating over ξ i from a i to ε i ,   i = 1 , 2 , , n , we observe that
a n ε n a 1 ε 1 i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n × a n ε n a 1 ε 1 i = 1 n h i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ 1 Δ ξ n = n i = 1 n 1 / p i i = 1 n 1 / p i n × i = 1 n h i a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i .
Applying (13) on a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i with f ( ξ i ) = 1 and
g ( ξ i ) = a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i ,
we have
a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i ε i a i 1 r i a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i ,
and then
i = 1 n a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i 1 p i Δ ξ i i = 1 n ε i a i 1 r i i = 1 n a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i .
Substituting (58) into (57), we have
a n ε n a 1 ε 1 i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n h i ε i a i 1 q i × i = 1 n a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i 1 p i .
Applying (12) on a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i with f ( ξ i ) = a i ξ i ( λ i σ ( t i ) h i 1   λ i Δ ( t i ) ) p i Δ t i and g Δ ( ξ i ) = 1 , we obtain
a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i = g ( ξ i ) a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i a i ε i a i ε i g σ ( ξ i ) λ i σ ( ξ i ) h i 1 λ i Δ ( ξ i ) p i Δ ξ i ,
where g ( ξ i ) = ξ i ε i . Since g ( ε i ) = 0 , we know from (60) that
a i ε i a i ξ i λ i σ ( t i ) h i 1 λ i Δ ( t i ) p i Δ t i Δ ξ i = a i ε i ε i σ ( ξ i ) λ i σ ( ξ i ) h i 1 λ i Δ ( ξ i ) p i Δ ξ i ,
Substituting (61) into (59), we observe that
a n ε n a 1 ε 1 i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n h i ε i a i 1 q i × i = 1 n a i ε i ε i σ ( ξ i ) λ i σ ( ξ i ) h i 1 λ i Δ ( ξ i ) p i Δ ξ i 1 p i .
From (50), the inequality (62) becomes
a n ε n a 1 ε 1 i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Δ ξ 1 Δ ξ n Q i = 1 n a i ε i ε i σ ( ξ i ) λ i σ ( ξ i ) h i 1 λ i Δ ( ξ i ) p i Δ ξ i 1 p i ,
which is (49). □
Remark 1.
If T = R and a i = 0 for i = 1 , 2 , , n , then we obtain the inequality (9) for the non-negative increasing function λ with λ i ( 0 ) = 0 ,   i = 1 , 2 , , n .
In the following remark, we present the discrete analogue of (9), i.e., when T = N , which is new and interesting for the reader.
Corollary 7.
If T = N , h i 1 ,   p i , r i > 1 such that 1 / p i + 1 / r i = 1 and λ i is a non-negative and increasing sequence with λ i ( a i ) = 0 ;   i = 1 , 2 , , n , then
ξ 1 = a 1 ε 1 1 ξ n = a n ε n 1 i = 1 n λ i h i ( ξ i ) i = 1 n ξ i a i / r i i = 1 n 1 / r i Q i = 1 n ξ i = a i ε i 1 ε i ξ i 1 λ i ( ξ i + 1 ) h i 1 Δ λ i ( ξ i ) p i 1 p i ,
where
Q = n i = 1 n 1 / p i i = 1 n 1 / p i n i = 1 n h i ε i a i 1 r i .

4. Conclusions and Future Work

In this paper, we establish some new dynamic Hilbert-type inequalities on time scale delta calculus by applying Hölder’s inequality, the chain rule and the mean inequality. In the future, we will prove Hilbert-type inequalities on diamond—α calculus and fractional conformable calculus.

Author Contributions

Software and writing—original draft, H.M.R. and A.I.S.; writing—review and editing, M.Z., A.A.I.A.-T. and M.A. All authors have read and agreed to the published version of the manuscript.

Funding

The authors extend their appreciation to the Deanship of Research and Graduate Studies at King Khalid University for funding this work through large Research Project under grant number RGP 2/190/45.

Data Availability Statement

Data are contained within the article.

Acknowledgments

The authors extend their appreciation to the Deanship of Research and Graduate Studies at King Khalid University for funding this work through large Research Project under grant number RGP 2/190/45.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Hilbert, D. Grundzüge Einer Allgemeinen Theorie der Linearen Intergraleichungen; Springer: Göttingen, Germany, 1906; pp. 157–227. [Google Scholar]
  2. Schur, I. Bernerkungen sur theorie der beschrankten Bilinearformen mit unendlich vielen veranderlichen. J. Math. 1911, 140, 1–28. [Google Scholar]
  3. Liu, Q.; Sun, W. A Hilbert-type fractal integral inequality and its applications. J. Inequal. Appl. 2017, 2017, 83. [Google Scholar] [CrossRef] [PubMed]
  4. Debnath, L.; Yang, B. Recent developments of Hilbert-type discrete and integral inequalities with applications. Int. J. Math. Math. Sci. 2012, 2012, 871845. [Google Scholar] [CrossRef]
  5. Guariglia, E. Riemann zeta fractional derivative-functional equation and link with primes. Adv. Differ. Equ. 2019, 2019, 261. [Google Scholar] [CrossRef]
  6. Hardy, G.H.; Littlewood, J.E.; Polya, G. The maximum of a certain bilinear form. Proc. Lond. Math. Soc. 1926, 25, 265–282. [Google Scholar] [CrossRef]
  7. Hardy, G.H. Notes on a theorem of Hilbert. Math. Z. 1920, 6, 314–317. [Google Scholar] [CrossRef]
  8. Hardy, G.H. Note on a theorem of Hilbert concerning series of positive term. Proc. Lond. Math. Soc. 1925, 23, 45–46. [Google Scholar]
  9. Hardy, G.H.; Littlewood, J.E.; Polya, G. Inequalities, 2nd ed.; Cambridge University Press: Cambridge, UK, 1934. [Google Scholar]
  10. Pachpatte, B.G. A note on Hilbert type inequality. Tamkang J. Math. 1998, 29, 293–298. [Google Scholar] [CrossRef]
  11. Pachpatte, B.G. Inequalities Similar to Certain Extensions of Hilbert’s Inequality. J. Math. Anal. Appl. 2000, 243, 217–227. [Google Scholar] [CrossRef]
  12. Kim, Y.H.; Kim, B.I. An Analogue of Hilbert’s inequality and its extensions. Bull. Korean Math. Soc. 2002, 39, 377–388. [Google Scholar] [CrossRef]
  13. Zhao, C.-J.; Chen, L.-Y.; Cheung, W.S. On some new Hilbert-type inequalities. Math. Slovaca 2011, 61, 15–28. [Google Scholar]
  14. Řehak, P. Hardy inequality on time scales and its application to half-linear dynamic equations. J. Inequal. Appl. 2005, 2005, 495–507. [Google Scholar] [CrossRef]
  15. Saker, S.H.; Ahmed, A.M.; Rezk, H.M.; O’Regan, D.; Agarwal, R.P. New Hilbert dynamic inequalities on time scales. Math. Inequal. Appl. 2017, 20, 1017–1039. [Google Scholar] [CrossRef]
  16. Ahmed, A.M.; AlNemer, G.; Zakarya, M.; Rezk, H.M. Some Dynamic Inequalities of Hilbert’s Type. J. Funct. Spaces 2020, 2020, 4976050. [Google Scholar] [CrossRef]
  17. Gulsen, T.; Jadlovská, I.; Yilmaz, E. On the number of eigenvalues for parameter-dependent diffusion problem on time scales. Math. Metho. Appl. Sci. 2021, 44, 985–992. [Google Scholar] [CrossRef]
  18. AlNemer, G.; Zakarya, M.; Abd El-Hamid, H.A.; Agarwal, P.; Rezk, H.M. Some Dynamic Hilbert-Type Inequalities on Time Scales. Symmetry 2020, 12, 1410. [Google Scholar] [CrossRef]
  19. El-Hamid, H.A.A.; Rezk, H.M.; Ahmed, A.M.; AlNemer, G.; Zakarya, M.; ELsaify, H.A. Some Dynamic Hilbert-Type Inequalities for Two Variables on Time Scales. J. Inequal. Appl. 2021, 2021, 31. [Google Scholar] [CrossRef]
  20. Bohner, M.; Peterson, A. Dynamic Equations on Time Scales: An Introduction with Applications; Birkhäuser: Boston, MA, USA, 2001. [Google Scholar]
  21. Agarwal, R.P.; O’Regan, D.; Saker, S.H. Dynamic Inequalities on Time Scales; Springer: Cham, Switzerland, 2014. [Google Scholar]
  22. Agarwal, R.P.; O’Regan, D.; Saker, S.H. Hardy Type Inequalities on Time Scales; Springer: Cham, Switzerland, 2016. [Google Scholar]
  23. Tuna, A.; Kutukcu, S. Some integral inequalities on time scales. Appl. Math. Mech. 2008, 29, 23–29. [Google Scholar] [CrossRef]
  24. Bohner, M.; Guseinov, G.S. Multiple integration on time scales. Dyn. Syst. Appl. 2005, 14, 579–606. [Google Scholar]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zakarya, M.; Saied, A.I.; Al-Thaqfan, A.A.I.; Ali, M.; Rezk, H.M. On Some New Dynamic Hilbert-Type Inequalities across Time Scales. Axioms 2024, 13, 475. https://doi.org/10.3390/axioms13070475

AMA Style

Zakarya M, Saied AI, Al-Thaqfan AAI, Ali M, Rezk HM. On Some New Dynamic Hilbert-Type Inequalities across Time Scales. Axioms. 2024; 13(7):475. https://doi.org/10.3390/axioms13070475

Chicago/Turabian Style

Zakarya, Mohammed, Ahmed I. Saied, Amirah Ayidh I Al-Thaqfan, Maha Ali, and Haytham M. Rezk. 2024. "On Some New Dynamic Hilbert-Type Inequalities across Time Scales" Axioms 13, no. 7: 475. https://doi.org/10.3390/axioms13070475

APA Style

Zakarya, M., Saied, A. I., Al-Thaqfan, A. A. I., Ali, M., & Rezk, H. M. (2024). On Some New Dynamic Hilbert-Type Inequalities across Time Scales. Axioms, 13(7), 475. https://doi.org/10.3390/axioms13070475

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop