Skip to main content

On some Volterra–Fredholm and Hermite–Hadamard-type fractional integral inequalities

Abstract

The main aim of this paper is establishing some new Volterra–Fredholm and Hermite–Hadamard-type fractional integral inequalities, which can be used as auxiliary tools in the study of solutions to fractional differential equations and fractional integral equations. Applications are also given to explicate the availability of our results.

Introduction

The subject of fractional calculus has gained considerable popularity and importance over the past few decades, mainly due to its validated applications in various fields of science and engineering [14]. Integral inequalities, especially fractional integral inequalities, have been paid more and more attention in recent years. These inequalities play important roles in the study of fractional differential equations and fractional integral equations. At present, many scholars are devoted to studying various integral inequalities, such as Volterra–Fredholm and Hermite–Hadamard-type inequalities. In [510] the authors generalized and analyzed the Volterra–Fredholm-type and delay integral inequalities. In addition, some applications in fractional differential equations were presented to illustrate the validity of their outcomes. Convex functions have found an important place in modern mathematics, as they can be seen in a large number of research papers and books today. In this context, the Hermite–Hadamard inequality can be regarded as the first fundamental result for convex functions, which is defined over an interval of real numbers with natural geometric interpretation and many applications. In [1120] a number of Hermite–Hadamard-type inequalities are deduced involving the classical and Riemann–Liouville fractional integrals for different classes of convex functions such as \((s,m)\)-convex, m-convex, log-convex, and prequasi-invex functions. In this paper, we consider Volterra–Fredholm and Hermite–Hadamard-type inequalities involving fractional integrals.

The structure of this paper is as follows. The first part gives some preliminary results about fractional integrals, derivatives, and convex functions. In the second part, we derive some new nonlinear Volterra–Fredholm-type fractional integral inequalities on time scales for one- and two-variable functions. In the third part, we establish Hermite–Hadamard-type inequalities and some other integral inequalities for the Riemann–Liouville fractional integral. Finally, we give some concluding remarks.

Preliminaries

In this section, we recall several definitions needed for the discussion.

Definition 2.1

([2123])

The Riemann–Liouville integral of a function \(f(x)\) of order \(\alpha >0\) is defined as

$$ {}_{\mathrm{RL}}{\mathrm{D}}^{-\alpha }_{a, t}f(t)= \frac{1}{\Gamma (\alpha )} \int _{a} ^{t} (t-u )^{ \alpha -1}f(u) \, \mathrm{d} u,\quad t>a>0, $$
(2.1)

where Γ is the gamma function.

Definition 2.2

([2123])

The Riemann–Liouville derivative of a function \(f(x)\) of order α is defined as

$$ \begin{aligned} {}_{\mathrm{RL}}{ \mathrm{D}}^{\alpha }_{a, t}f(t)&= \frac{ \mathrm{d}^{n}}{\mathrm{d}t^{n}} \bigl({}_{\mathrm{RL}}{\mathrm{D}}^{-(n-\alpha )}_{a, t}f(t) \bigr) \\ &=\frac{1}{\Gamma (n- \alpha )}\frac{ \mathrm{d}^{n}}{\mathrm{d}t^{n}} \int _{a}^{t} (t-u )^{n - \alpha -1}f(u) \, \mathrm{d}u, \quad t>a>0, \end{aligned} $$
(2.2)

where \(n-1<\alpha <n\in \mathbb{Z}^{+}\).

Definition 2.3

([2123])

The Caputo derivative of a function \(f(x)\) of order α is defined as

$$ \begin{aligned} {}_{C}{ \mathrm{D}}^{\alpha }_{a, t}f(t)&={{}_{\mathrm{RL}}{ \mathrm{D}}^{-(n- \alpha )}_{a, t} f^{(n)}(t)} \\ &=\frac{1}{\Gamma (n- \alpha )} \int _{a}^{t} (t-u )^{n- \alpha -1}f^{(n)}(u) \,\mathrm{d}u,\quad t>a, \end{aligned} $$
(2.3)

where \(n-1<\alpha <n\in \mathbb{Z}^{+}\).

Definition 2.4

([11, 12])

A function \(f:[a, b]\subset \mathbb{R}\mapsto \mathbb{R}\) is said to be convex on \([a, b]\) if

$$ f \bigl(\lambda x+(1-\lambda ) y \bigr)\leq \lambda f(x)+(1- \lambda )f(y), $$
(2.4)

for all \(x,y\in [a, b]\) and \(\lambda \in [0, 1]\).

Definition 2.5

([11])

A function \(f:[a, b]\subset \mathbb{R}\mapsto \mathbb{R}\) is said to be \((s, m)\)-convex with modulus \(\mu \geq 0\) (in the second sense) if

$$ f \bigl(\lambda x+m(1-\lambda ) y \bigr)\leq \lambda ^{s} f(x) +m(1- \lambda )^{s}f(y)- \mu t(1-t) (x-y)^{2} $$
(2.5)

for all \(x,y\in [a, b]\), \(\lambda \in [0, 1]\), and \(s,m\in [0, 1]\).

Definition 2.6

([18, 24])

A function \(f:[a, b]\subset \mathbb{R}\mapsto \mathbb{R}\) is said to be s-convex (in the second sense) if

$$ f \bigl(\lambda x+(1-\lambda ) y \bigr)\leq \lambda ^{s} f(x) +(1- \lambda )^{s}f(y) $$
(2.6)

for all \(x, y\in [a, b]\), \(\lambda \in [0, 1]\), and \(s\in [0, 1]\).

Definition 2.7

([25])

Let \(f:[a, b]\subset \mathbb{R}\mapsto \mathbb{R}\) be a convex function. Then the Hermite–Hadamard inequality is given by

$$ f \biggl(\frac{a+b}{2} \biggr)\leq \frac{1}{b -a} \int ^{b}_{a}f(t) \,\mathrm{d}t\leq \frac{f(a)+f(b)}{2}. $$
(2.7)

Nonlinear Volterra–Fredholm-type fractional integral inequalities

In this section, we show and prove certain Riemann–Liouville fractional integral inequalities of nonlinear Volterra–Fredholm-type by amplification, differentiation, integration, and inverse functions.

In the following discussion, we assume that

  1. 1.

    \(x(t), f(t), c(t), g_{1}(t), g_{2}(t),r(t)\in C([a, b], \mathbb{R}\mathbbm{_{+}})\) with \(r(t)\leq t\),

  2. 2.

    \(h(t)\in C(\mathbb{R}\mathbbm{_{+}}, \mathbb{R}\mathbbm{_{+}})\) is a nondecreasing function with \(h(t)\geq 1\),

  3. 3.

    \(Q(v)=\int _{0}^{v}\frac{\mathrm{d}r}{h(r)}\) for \(v\geq 0\).

However, for fractional order α, we only consider the case \(1<\alpha <2\).

Theorem 3.1

If

$$ x(t)\leq (t-a)c(t)+ \int _{a}^{t}(t-u)^{\alpha -1} f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u, $$
(3.1)

then

$$ x(t)\leq \biggl(\frac{t-a}{b-a} \biggr)^{\alpha -1} Q^{-1} \biggl[(b-a)^{\alpha -1}(t-a)^{2- \alpha }c(t)+ (b-a)^{\alpha -1} \int _{a}^{t}f(s)\,\mathrm{d}s \biggr]. $$
(3.2)

Proof

From (3.1) it follows that

$$ \begin{aligned} (t-a)^{1-\alpha }x(t)&\leq (t-a)^{2-\alpha }c(t)+(t-a) ^{1- \alpha } \int _{a}^{t}(t -u)^{\alpha -1} f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\leq (t-a)^{2-\alpha } c(t)+ \int _{a}^{t}f(u)h \bigl(x (u) \bigr)\, \mathrm{d}u. \end{aligned} $$
(3.3)

For \(t\in [a,b]\), defining the function \(y(t)\) by the right-hand side of (3.3), we have

$$ \begin{aligned} y'(t)&= \bigl[(t-a)^{2-\alpha }c(t) \bigr]' +f(t)h \bigl(x(t) \bigr) \\ &\leq \bigl[(t-a)^{2-\alpha }c(t) \bigr]' +f(t)h \bigl((t-a)^{\alpha -1 }y(t) \bigr) \\ &\leq \bigl[(t-a)^{2-\alpha }c(t) \bigr]' +f(t)h \bigl((b-a)^{\alpha -1 }y(t) \bigr), \end{aligned} $$

which implies

$$ \begin{aligned} \frac{y'(t)}{h((b-a) ^{\alpha -1}y(t))} &\leq \frac{[(t-a)^{2 -\alpha }c(t)]'}{h((b -a)^{\alpha -1}y(t))} +f(t) \\ &\leq \bigl[(t-a)^{2-\alpha }c(t) \bigr]' +f(t). \end{aligned} $$
(3.4)

Multiplying both sides of (3.4) by \((b-a)^{\alpha -1}\), we get

$$ \frac{(b-a)^{\alpha -1} y'(t)}{h((b-a)^{\alpha -1}y(t))}\leq (b-a)^{ \alpha -1} \bigl[(t-a)^{2- \alpha }c(t) \bigr]'+(b-a) ^{\alpha -1}f(t). $$
(3.5)

Setting \(t=s\) and integrating both sides of (3.5) over \([a, t]\), we find

$$ Q \bigl((b-a)^{\alpha -1}y(t) \bigr) \leq (b-a)^{\alpha -1}(t- a)^{2-\alpha }c(t)+(b-a) ^{\alpha -1} \int _{a}^{t} f(s)\,\mathrm{d}s. $$

In fact, \(Q^{-1}\) is nondecreasing, and \(y(a)=0\). We can deduce that

$$ y(t)\leq (b-a)^{1-\alpha } Q^{-1} \biggl[(b-a)^{ \alpha -1}(t-a)^{2- \alpha }c(t)+(b-a)^{\alpha -1} \int _{a}^{t}f(s)\,\mathrm{d}s \biggr]. $$
(3.6)

Using (3.3) and (3.6), we can derive the desired inequality (3.2). This ends the proof. □

Theorem 3.2

If \(x(t)\) satisfies

$$ \begin{aligned} x(t)&\leq (t-a)c(t)+ \int _{a}^{t}(t-u)^{\alpha -1}g_{1}(u) f(u)h \bigl(x(u) \bigr)\,\mathrm{d}u \\ &\quad {} + \int _{a}^{t}(t-u)^{\alpha -1}g_{1}(u) \int _{a}^{u}(u-v) ^{ \alpha -1}g_{2} (v)h \bigl(x(v) \bigr)\,\mathrm{d}v \,\mathrm{d}u, \end{aligned} $$
(3.7)

then

$$ \begin{aligned} x(t)&\leq \biggl( \frac{ t-a}{b-a} \biggr)^{\alpha -1}Q^{-1} \biggl\{ (b-a)^{\alpha -1}(t-a)^{2- \alpha }c(t)+(b-a)^{ \alpha -1} \\ &\quad {} \times \int _{a}^{t}g_{1}(s) \biggl[ f(s)+2^{2- \alpha }t^{ \alpha -1} \int _{a}^{s}g_{2}(v) \, \mathrm{d}v- \int _{a} ^{s}v^{\alpha -1}g_{2} (v)\,\mathrm{d}v \biggr] {\, \mathrm{d}}s \biggr\} . \end{aligned} $$
(3.8)

Proof

Simplifying (3.7), we can easily get that

$$ \begin{aligned} (t-a)^{1-\alpha }x(t) &\leq (t-a)^{2-\alpha } c(t)+ \int _{a}^{t}g_{1} (u)f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\quad {} + \int _{a}^{t} g_{1}(u) \int _{a}^{u} (u-v)^{\alpha -1}g_{2} (v)h \bigl(x(v) \bigr) \,\mathrm{d} v \,\mathrm{d}u. \end{aligned} $$
(3.9)

It is known that \((u-v)^{\alpha -1}\leq 2^{2-\alpha } u^{\alpha -1}-v^{\alpha -1}\). Therefore

$$ \begin{aligned} (t-a)^{1-\alpha }x(t) &\leq (t-a)^{2-\alpha } c(t)+ \int _{a}^{t}g_{1} (u)f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\quad {} + \int _{a}^{t} g_{1}(u) \int _{a}^{u} \bigl(2^{2-\alpha }u^{ \alpha -1}-v^{\alpha -1} \bigr)g_{2}(v)h \bigl(x(v) \bigr)\,\mathrm{d}v \,\mathrm{d}u. \end{aligned} $$
(3.10)

For the convenience of calculation, denote the right-hand side of (3.10) as \(y(t)\). Then

$$ \begin{aligned} y'(t) &= \bigl[(t-a)^{2- \alpha }c(t) \bigr]'-g_{1}(t) \int _{a}^{t}v^{ \alpha -1}g_{2}(v)h \bigl(x(v) \bigr)\,\mathrm{d}v \\ &\quad {} +g_{1} (t) \biggl[ f(t)h \bigl(x(t) \bigr) +2^{2-\alpha }t^{\alpha -1} \int _{a}^{t}g_{2} (v)h \bigl(x(v) \bigr)\,\mathrm{d} v \biggr] \\ &\leq \bigl[(t-a)^{2-\alpha } c(t) \bigr]'+g_{1}(t)h \bigl((t-a) ^{\alpha -1}y(t) \bigr) \biggl[ f(t)+2^{2-\alpha }t^{ \alpha -1} \int _{a}^{t} g_{2}(v) \, \mathrm{d}v \biggr] \\ &\quad {} -g_{1}(t)h \bigl((t-a) ^{\alpha -1}y(t) \bigr) \int _{a}^{t}v^{\alpha -1}g_{ 2}(v) \,\mathrm{d}v. \end{aligned} $$
(3.11)

Hence

$$ \begin{aligned} y'(t) &\leq \bigl[(t-a)^{2-\alpha } c(t) \bigr]'+g_{1}(t)h \bigl((b-a) ^{ \alpha -1}y(t) \bigr) \biggl[ f(t)+2^{2-\alpha }t^{ \alpha -1} \int _{a}^{t} g_{2}(v) \, \mathrm{d}v \biggr] \\ & \quad {}-g_{1}(t)h \bigl((t-a) ^{\alpha -1}y(t) \bigr) \int _{a}^{t}v^{\alpha -1}g_{ 2}(v) \,\mathrm{d}v. \end{aligned} $$
(3.12)

By the same steps from (3.4)–(3.6), as in the proof of Theorem 3.1, we have

$$ \begin{aligned} y(t) &\leq (b-a)^{1-\alpha } Q^{-1} \biggl\{ (t-a)^{2 - \alpha }c(t)+(b-a)^{ \alpha -1} \int _{a}^{t} g_{1}(s)f(s)+2^{2- \alpha }t^{ \alpha -1} \\ &\quad {} \times \biggl[ \int _{a}^{s}g_{2}(v)\, \mathrm{d}v- \int _{a}^{s} v^{\alpha -1}g_{2}(v) \,\mathrm{d}v \biggr]{\,\mathrm{d}}s \biggr\} . \end{aligned} $$
(3.13)

Combining (3.10) and (3.13), we can easily find (3.8). The proof is completed. □

Theorem 3.3

Let \(S(t)=Q(2t+c(b)-2c(a))-Q(t)\) be a nondecreasing function. If there is a function \(x(t)\) such that

$$ \begin{aligned} x(t)&\leq c(t)+ \int _{r(a)}^{r(t)} \bigl(r(t) -u \bigr)^{\alpha -1}g_{1} (u)f(u)h \bigl(x(u) \bigr)\,\mathrm{d} u \\ &\quad {} + \int _{r(a)}^{r(b)} \bigl(r(b)-u \bigr)^{\alpha -1}g_{1}(u) f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r(t)} \bigl(r(t) -u \bigr)^{\alpha -1}g_{1} (u) \int _{r(a)} ^{u}(u-v)^{\alpha -1} g_{2}(v)h \bigl(x(v) \bigr)\,\mathrm{d}v\,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r(b)} \bigl(r(b)-u \bigr)^{\alpha -1}g_{1} (u) \int _{r(a)}^{u}(u-v)^{\alpha -1}g_{2}(v)h \bigl(x(v) \bigr) \,\mathrm{d}v\,\mathrm{d}u, \end{aligned} $$
(3.14)

then

$$\begin{aligned} x(t)&\leq Q^{-1} \biggl\{ c(t)-c(a)+Q \biggl[S^{-1} \biggl(c(b)-c(a)+ \bigl(r(b)- r(a) \bigr)^{\alpha -1} \\ &\quad {} \times \int _{r(a)}^{r(b)}g_{1} (s) \biggl(f(s)+ \int _{r(a)} ^{s} \bigl(2^{2-\alpha }r^{\alpha -1}(s)-v^{\alpha -1} \bigr)g_{2}(v) \,\mathrm{d} v \biggr)\,\mathrm{d}s \biggr) \biggr] \\ &\quad {} + \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{r(t)}g_{1} (s)f(s) \, \mathrm{d}s+ \bigl(r(b)-r (a) \bigr)^{\alpha -1} \\ &\quad {} \times \int _{r(a)}^{r (t)}g_{1}(s) \int _{r( a)}^{s} \bigl(2^{2- \alpha }r^{ \alpha -1}(s)-v^{\alpha -1} \bigr) g_{2}(v)\,\mathrm{d}v \,\mathrm{d}s \biggr\} . \end{aligned}$$
(3.15)

Proof

According to (3.14), we have

$$ \begin{aligned} x(t)&\leq c(t)+ \bigl(r(b)-r(a ) \bigr)^{\alpha -1} \int _{r(a)}^{r(t)}g_{1}(u)f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\quad {} + \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{r(b)}g_{1} (u)f(u)h \bigl(x(u) \bigr) \,\mathrm{d}u \\ &\quad {} + \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{r(t)}g_{1} (u) \int _{r(a)}^{u}(u-v)^{ \alpha -1}g_{2}(v)h(x(v) \,\mathrm{d}v\,\mathrm{d}u \\ &\quad {} + \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{r(b)}g_{1} (u) \int _{r(a)}^{u}(u-v)^{ \alpha -1}g_{2}(v)h \bigl(x(v) \bigr) \,\mathrm{d}v\,\mathrm{d}u. \end{aligned} $$
(3.16)

Denote the right-hand side of (3.16) as \(y(t)\). Inspired by (3.9)–(3.11), \(y(t)\) has the following estimate:

$$ \begin{aligned} y(t)&\leq Q^{-1} \biggl\{ Q \bigl(y(a) \bigr)+c(t)-c(a)+ \bigl(r(b) -r(a) \bigr)^{ \alpha -1} \\ &\quad {} \times \int _{r(a)}^{r(t)}g_{1}(s) \biggl[ f(s)+ \int _{r(a)} ^{s} \bigl(2^{2-\alpha }r^{ \alpha -1}(s)-v^{\alpha -1} \bigr)g_{2}(v) \,\mathrm{d} v \biggr]\,\mathrm{d}s \biggr\} . \end{aligned} $$
(3.17)

By the definition of \(y(t)\) we get

$$ \begin{aligned} y(b) &=c(b)+2 \bigl(r(b)-r(a) \bigr)^{ \alpha -1} \int _{r(a)}^{r(b)}g_{1}(u)f(u)h \bigl(x (u) \bigr)\,\mathrm{d}u \\ &\quad {} +2 \bigl(r(b)-r(a) \bigr)^{ \alpha -1} \int _{r(a)}^{r(b)}g_{1}(u) \int _{r (a)}^{u}(u-v)^{ \alpha -1}g_{2}(v)h \bigl(x(v) \bigr) \,\mathrm{d}v\,\mathrm{d}u \\ &=c(b)+2 \bigl(y(a)-c(a) \bigr). \end{aligned} $$
(3.18)

According to (3.17), we have

$$ \begin{aligned}Q \bigl(y(b) \bigr) &\leq Q \bigl(y(a) \bigr)+c(b)- c(a)+ \bigl(r(b)-r(a) \bigr)^{ \alpha -1} \\ &\quad {} \times \int _{r(a)} ^{r(b)}g_{1}(s) \biggl[ f(s)+ \int _{r(a)}^{s} \bigl(2^{2-\alpha }r^{ \alpha -1}(s)-v^{\alpha -1} \bigr)g_{2}(v) \,\mathrm{d}v \biggr] \,\mathrm{d}s. \end{aligned} $$
(3.19)

Thus

$$ \begin{aligned} S \bigl(y(a) \bigr)&=Q \bigl(2y(a)+c(b)- 2c(a) \bigr)-Q \bigl(y(a) \bigr) \\ &\leq c(b)-c(a)+ \bigl(r(b) -r(a) \bigr)^{\alpha -1} \\ &\quad {} \times \int _{r(a)} ^{r(b)}g_{1}(s) \biggl[f (s)+ \int _{r(a)}^{s} \bigl(2^{2-\alpha }r^{\alpha -1} (s)-v^{\alpha -1} \bigr)g_{2}(v) \,\mathrm{d}v \biggr] \, \mathrm{d}s. \end{aligned} $$
(3.20)

Since Q and S are nondecreasing, we have

$$ \begin{aligned} & Q \bigl(y(a) \bigr) \\ &\quad \leq Q \biggl\{ S^{-1} \biggl[c(b)-c(a)+ \bigl(r(b) -r(a) \bigr)^{\alpha -1} \\ &\qquad {} \times \int _{r(a)}^{r(b)}g_{1} (s) \biggl(f(s)+ \int _{r( a)}^{s} \bigl(2^{2-\alpha } r^{\alpha -1}(s)-v^{\alpha -1} \bigr)g_{2}(v)\,\mathrm{d}v \biggr)\,\mathrm{d}s \biggr] \biggr\} . \end{aligned} $$
(3.21)

Using (3.17) and (3.21), it follows that

$$ \begin{aligned} y(t)&\leq Q^{-1} \biggl\{ c(t)-c(a)+Q \biggl[S^{-1} \biggl(c(b)-c(a)+ \bigl(r(b)-r( a) \bigr)^{\alpha -1} \\ &\quad {} \times \int _{r(a)}^{r(b)} g_{1}(s) \biggl(f(s)+ \int _{r(a)}^{s} \bigl(2^{2- \alpha }r^{\alpha -1}(s)- v^{\alpha -1} \bigr)g_{2} (v)\,\mathrm{d}v \biggr) \, \mathrm{d}s \biggr) \biggr] \\ &\quad {} + \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{s} \bigl(2^{2-\alpha }r^{ \alpha -1}(s) -v^{\alpha -1} \bigr)g_{2} (v)\,\mathrm{d}v\,\mathrm{d}s \\ &\quad {} + \bigl(r(b)-r(a) \bigr) ^{\alpha -1} \int _{r(a)}^{r(t)}g_{1}(s)f(s) \, \mathrm{d}s \biggr\} . \end{aligned} $$
(3.22)

From (3.16) and (3.22) the expected result follows. □

Similarly to the above case with single-variable functions, we will consider bivariate functions.

Let \(I_{1}=[u_{0}, u_{T}]\) and \(I_{2}=[v_{0}, v_{T}]\) with \(u_{0}, v_{0} \geq 0\). Assume that:

  1. 1.

    \(x(u, v), f(u, v), c(u, v), g_{1}(u, v), g_{2} (u, v)\in C(I_{1} \times I_{2}, \mathbb{R}\mathbbm{_{+}})\),

  2. 2.

    \(r_{1}(u)\in C(I_{1}, \mathbb{R}\mathbbm{_{+}})\) and \(r_{2}(v)\in C(I_{2}, \mathbb{R}\mathbbm{_{+}})\) with \(r_{1}(u)\leq u\), \(r_{2}(v)\leq v\),

  3. 3.

    \(h(t)\in C(\mathbb{R}\mathbbm{_{+}}, \mathbb{R}\mathbbm{_{+}})\) is a nondecreasing function with \(h(t)\geq 1\),

  4. 4.

    \(Q(v)=\int _{0}^{v}\frac{\mathrm{d}r}{h(r)}\) for \(v\geq 0\).

Under such conditions, we state the following theorem.

Theorem 3.4

Suppose that \(x(u,v)\) satisfies the following inequality:

$$ \begin{aligned} & x(u,v) \\ &\quad \leq \bigl(r_{1}(u)-r_{1} (u_{0}) \bigr) \bigl(r_{2}(v)-r_{2}(v_{0}) \bigr)c(u,v) \\ &\qquad {} + \int ^{r_{1}(u)}_{r_{1}(u_{0})} \int ^{r_{2}(v)} _{r_{2}(v_{0})} \bigl(r_{1}(u)-s \bigr) ^{\alpha -1} \bigl(r_{2}(v)-\tau \bigr) ^{\alpha -1}g_{1}(s,\tau ) \biggl[f(s, \tau )h \bigl(x(s,\tau ) \bigr) \\ &\qquad {} + \int ^{s}_{r_{1} (u_{0})} \int ^{\tau }_{r_{2}(v _{0})}(s-w)^{ \alpha -1}(\tau - z)^{\alpha -1}g_{2}(w,z)h \bigl(x(w, z) \bigr)\,\mathrm{d}w \, \mathrm{d}z \biggr]\,\mathrm{d}s\,\mathrm{d} \tau \end{aligned} $$
(3.23)

for \(u\in I_{1}\), \(v\in I_{2}\). Then

$$\begin{aligned} x(u,v)&\leq \frac{1}{k} \bigl(u- r_{1}(u_{0}) \bigr)^{\alpha -1} \bigl( v-r_{2}(v_{0}) \bigr)^{ \alpha -1} \\ &\quad {} \times Q^{-1} \biggl\{ k \bigl(u-r_{1}(u_{0}) \bigr)^{2- \alpha } \bigl(v-r_{2}(v_{0}) \bigr)^{2-\alpha }c( u,v)-k_{0} k c(u_{0}, v) \\ &\quad {} +k \int _{r_{1}(u_{0}) }^{r_{1}(u)} \biggl[ \int ^{r _{2}(v)}_{r_{2}(v_{0})}g_{1}(s, \tau ) \biggl(f(s, \tau )+ \int ^{s}_{r_{1}(u_{ 0})} \int ^{ \tau }_{r_{2}(v_{0})}l(w,z,s,\tau ) \\ &\quad {} \times g_{2}(w,z)\,\mathrm{d}w\,\mathrm{d}z \biggr) \,\mathrm{d} \tau \biggr] \,\mathrm{d}s \biggr\} , \end{aligned}$$
(3.24)

where \(l(w,z,s,\tau )=(2^{2-\alpha }s^{\alpha -1}-w^{\alpha -1})(2^{2-\alpha } \tau ^{\alpha -1}- z^{\alpha -1})\). The constants \(k_{0}\) and k are defined by \(k_{0}=(u-r_{1}(u_{0}))^{2-\alpha }(v -r_{2}(v_{0}))^{2-\alpha }\) and \(k=(u_{T}-r_{1}( u_{0}))^{\alpha -1}(v_{T}-r_{2}(v_{0}))^{\alpha -1 }\), respectively.

Proof

From (3.23) we easily derive that

$$ \bigl(u-r_{1}(u_{0}) \bigr)^{1-\alpha } \bigl(v-r_{ 2}(v_{0}) \bigr)^{1-\alpha }x(u,v)\leq y(u,v), $$
(3.25)

where

$$ \begin{aligned} & y(u,v) \\ &\quad = \bigl(u-r_{1}(u_{0}) \bigr)^{2-\alpha } \bigl(v-r_{2}(v_{0}) \bigr)^{2-\alpha }c (u,v) \\ &\qquad {} + \int ^{r_{1}(u)} _{r_{1}(u_{0})} \int ^{r_{2} (v)}_{r_{2}(v_{0})}g_{1}(s, \tau )f(s, \tau )h \bigl(x(s, \tau ) \bigr)\,\mathrm{d}s\,\mathrm{d} \tau \\ &\qquad {} + \int ^{r_{1}(u)} _{r_{1}(u_{0})} \int ^{r_{2} (v)}_{r_{2}(v_{0})}g_{1}(s, \tau ) \int ^{s}_{r_{1} (u_{0})} \int ^{\tau }_{r_{2} (v_{0})}l(w,z,s, \tau )g_{2} (w,z)h \bigl(x(w,z) \bigr)\,\mathrm{d}w \,\mathrm{d}z\,\mathrm{d}s \,\mathrm{d} \tau . \end{aligned} $$
(3.26)

Taking the partial derivative of \(y(u,v)\) with respect to u, we have

$$\begin{aligned} \frac{\partial y(u,v)}{ \partial u} &= \frac{\partial (u-r_{1}(u_{0})^{2-\alpha }(v-r _{2}(v_{0}))^{2-\alpha }c(u, v) )}{\partial u} \\ &\quad {} +r'_{1}(u) \int ^{r_{2} (v)}_{r_{2}(v_{0})}g_{1} \bigl(r _{1}(u), \tau \bigr)f \bigl(r_{1} (u), \tau \bigr)h \bigl(x \bigl(r_{1}(u), \tau \bigr) \bigr)\,\mathrm{d}\tau \\ &\quad {} +r'_{1}(u) \int ^{r_{2}(v)}_{r_{2}(v_{0})}g_{1} \bigl(r_{1}(u), \tau \bigr) \int ^{r_{1}(u)}_{r_{1}(u_{0}) } \int ^{\tau }_{r_{2}(v_{0} )}l(w,z,s, \tau ) \\ &\quad {} \times g_{2} (w,z)h \bigl(x(w,z) \bigr)\, \mathrm{d}w\, \mathrm{d}z \,\mathrm{d}\tau \\ &\leq \frac{\partial (u -r_{1}(u_{0}))^{2-\alpha }( v-r_{2}(v_{0}))^{2-\alpha } c(u,v) )}{\partial u} \\ &\quad {} +r'_{1}(u)h \bigl( \bigl(u_{T} -r_{1}(u_{0}) \bigr)^{\alpha -1} \bigl(v_{T}-r_{2}(v_{0}) \bigr)^{ \alpha -1}y(u,v) \bigr) \\ &\quad {} \times \int ^{r_{2}(v)} _{r_{2}(v_{0})}g_{1} \bigl(r_{1} (u),\tau \bigr) \biggl[f \bigl(r_{1}(u), \tau \bigr)+ \int ^{r_{1}(u)}_{r_{ 1}(u_{0})} \int ^{\tau }_{r_{ 2}(v_{0})} l(w,z,s,\tau ) g_{2}(w,z)\,\mathrm{d}w \,\mathrm{d}z \biggr] \, \mathrm{d}\tau . \end{aligned}$$
(3.27)

Through a series of calculations, we get

$$ \begin{aligned} \frac{k\partial y(u, v)}{h(ky(u,v))\partial u}&\leq \frac{k\partial ((u-r_{1}(u_{0})) ^{2-\alpha }(v-r_{2}(v_{0}))^{2-\alpha }c(u,v) )}{\partial u} \\ &\quad {} +kr'_{1}(u) \int ^{r_{2}(v)}_{r_{2}(v_{0 })}g_{1} \bigl(r_{1}(u), \tau \bigr) f \bigl(r_{1}(u),\tau \bigr) \,\mathrm{d}\tau \\ &\quad {} +kr'_{1}(u) \int ^{r_{1}(u)}_{r_{1}(u_{0} )} \int ^{\tau }_{r_{2}(v_{0})}l(w,z,s,\tau )g_{2} (w,z)\,\mathrm{d}w\,\mathrm{d}z\,\mathrm{d}\tau , \end{aligned} $$
(3.28)

where

$$ k= \bigl(u_{T}-r_{1}(u_{0}) \bigr) ^{\alpha -1} \bigl(v_{T}-r_{2 }(v_{0}) \bigr)^{ \alpha -1}. $$

Integrating both sides of (3.28) with respect to t over \([u_{0}, u]\) yields the relation

$$ \begin{aligned} & Q \bigl( ky(u,v) \bigr) \\ &\quad \leq k \bigl(u-r_{1}(u_{0}) \bigr)^{2-\alpha } \bigl(v-r_{2}(v_{0}) \bigr)^{2-\alpha }c(u,v) \\ &\qquad {} -k_{0}kc(u_{0},v) +k \int _{r_{1}(u_{0})}^{r_{1}(u)} \int ^{r_{2}( v)}_{r_{2}(v_{0})}g_{1} (s,\tau )f(s, \tau )\,\mathrm{d}\tau \,\mathrm{d}s \\ &\qquad {} +k \int _{r_{1}(u_{0})}^{r_{1}(u)} \int ^{ r_{2}(v)}_{r_{2}(v_{0} )}g_{1}(s,\tau ) \int ^{s} _{r_{1}(u_{0})} \int ^{ \tau }_{r_{2}(v_{0})}l( w,z,s,\tau )g_{2}(w,z) \,\mathrm{d}w\,\mathrm{d} z\,\mathrm{d}\tau \, \mathrm{d}s, \end{aligned} $$
(3.29)

where

$$ k_{0}= \bigl(u-r_{1}(u_{0}) \bigr) ^{2-\alpha } \bigl(v-r_{2}( v_{0}) \bigr)^{2- \alpha }. $$

Since \(Q^{-1}\) is an increasing function, in the light of (3.25) and (3.29), we observe that (3.24) holds. The theorem is proved. □

To illustrate our results, the following Volterra–Fredholm fractional integral equations for one and two variables are separately considered in Corollaries 3.13.3:

$$\begin{aligned}& x(t)=A(t)+ \int _{a}^{t} (t-u)^{\alpha -1}g_{1}(u) \biggl[ f(u)x(u)+ \int _{a} ^{u}(u-v)^{\alpha -1}g_{2 }(v)x(v) \,\mathrm{d}v \biggr] \,\mathrm{d}u, \end{aligned}$$
(3.30)
$$\begin{aligned}& \begin{aligned} x(t)&=A(t)+ \int _{r(a)}^{ r(t)} \bigl(r(t)-u \bigr)^{\alpha -1} g_{1}(u)f(u) \sum _{i=1}^{ i=n} \gamma _{i}x^{\beta _{ i}}(u) \,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r(t) } \bigl(r(t)-u \bigr)^{\alpha -1}g_{ 1}(u) \int _{r(a)}^{u}( u-v)^{\alpha -1}g_{2}(v) \sum_{i=1}^{i=n}\gamma _{i}x^{\beta _{i}}(v) \,\mathrm{d}v\,\mathrm{d}u,\quad \beta _{i}>0, \end{aligned} \end{aligned}$$
(3.31)

and

$$ x(u,v)=C(u, v)+ \int _{ u_{0}}^{u} \int _{v_{0}}^{v}(u-s)^{ \alpha -1}(v- \tau )^{\alpha -1}F \bigl(s, \tau , x(s, \tau ),G(s, \tau ) \bigr)\,\mathrm{d}s \,\mathrm{d}\tau , $$
(3.32)

where \(G(s,\tau )=\int _{u _{0}}^{s}\int _{v_{0}}^{ \tau }(s-w)^{\alpha -1}( \tau -z)^{\alpha -1}x^{p} (w,z)\,\mathrm{d}w\,\mathrm{d}z\), \(0< p<1\).

Corollary 3.1

Suppose that \(x(t)\) satisfies

$$ \begin{aligned} & x(t) \\ &\quad \leq c(t)+ \int _{r(a)}^{r(t)} \bigl(r(t)-u \bigr) ^{\alpha -1}g_{1}(u) \biggl[ f(u)x(u)+ \int _{r(a)}^{u}(u-v)^{ \alpha -1}g_{2}(v)x (v)\,\mathrm{d}v \biggr] \,\mathrm{d}u \\ &\qquad {} + \int _{r(a)}^{ r(b)} \bigl(r(b)-u \bigr)^{\alpha -1}g_{1}(u) \biggl[f(u) x(u)+ \int _{r(a)}^{u} (u-v)^{\alpha -1}g_{2} (v)x(v)\,\mathrm{d}v \biggr] \,\mathrm{d}u \end{aligned} $$
(3.33)

for \(c(b)\leq 2c(a)+1\). Then we can get an explicit estimation of \(x(t)\) in (3.30):

$$ \begin{aligned} x(t)&\leq \frac{c(b)- (2c(a)+1)}{M-2} \biggl\{ \exp \biggl[c(t)-c(a) + \bigl(r(b)-r(a) \bigr)^{\alpha - 1} \\ &\quad {} \times \int _{r(a)}^{r(t)}g_{1} (s) \biggl(f(s)+ \int _{r (a)}^{s} \bigl(2^{2-\alpha }r ^{\alpha -1}(s)-v^{\alpha -1} \bigr)g_{2}(v) \,\mathrm{d}v \biggr)\,\mathrm{d}s \biggr]-1 \biggr\} , \end{aligned} $$
(3.34)

where

$$ \begin{aligned} M&=\exp \biggl\{ c(b)-c(a) + \bigl(r(b)-r(a) \bigr)^{\alpha -1 } \int _{r(a)}^{r(b)}g_{ 1}(s) \\ &\quad {} \times \biggl[f(s)+ \int _{r(a)}^{s} \bigl(2 ^{2-\alpha }r^{ \alpha -1} (s)-v^{\alpha -1} \bigr)g_{2} (v)\,\mathrm{d}v \biggr] \, \mathrm{d}s \biggr\} . \end{aligned} $$

Proof

Since \(x(t)\leq x(t)+1\), using \(h(t)=t+1\) in Theorem 3.3, we can get that \(Q(v)=\log (v+1)\), \(Q^{-1}(t)=\exp (t)-1\), \(S(t)=\log (2t+c(b)-2c(a)+1)- \log (t+1)\), and \(S^{-1}(t)=\frac{c(b)-(2c(a)+1)}{\exp (t)- 2}-1\). So (3.34) can be easily proved. □

Corollary 3.2

If \(r(t)\), a, b, and \(x(t)\) in (3.31) meet the \(r(t)\leq t\), \(1\leq a\), \(b \leq \log (2)+2c(a)-c(b)\), and

$$ \begin{aligned} x(t)&\leq c(t)+ \int _{ r(a)}^{r(t)} \bigl(r(t)-u \bigr)^{ \alpha -1}g_{1}(u)f(u) \sum_{i=1}^{i=n} \gamma _{i}x^{\beta _{i}}(u) \,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r(b) } \bigl(r(b)-u \bigr)^{\alpha -1}g_{ 1}(u)f(u) \sum _{i=1}^{i= n}\gamma _{i}x^{\beta _{i} }(u) \,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r (t)} \bigl(r(t)-u \bigr)^{\alpha -1} g_{1}(u) \int _{r(a)}^{u} (u-v)^{\alpha -1}g_{2}(v) \sum_{i=1}^{i=n}\gamma _{i}x^{\beta _{i}}(v) \,\mathrm{d}v\,\mathrm{d}u \\ &\quad {} + \int _{r(a)}^{r(b) } \bigl(r(b)-u \bigr)^{\alpha -1}g_{ 1}(u) \int _{r(a)}^{u}(u-v )^{\alpha -1}g_{2}(v) \sum_{i=1}^{i=n}\gamma _{i}x^{\beta _{i}}(v) \,\mathrm{d} v\,\mathrm{d}u, \end{aligned} $$
(3.35)

then

$$ x(t)\leq \log \biggl(\frac{1}{1-B-K(t)} \biggr), $$
(3.36)

where

$$ \begin{aligned} K(t)&=c(t)-c(a)+ \Biggl(\sum ^{i=n}_{i=1}\gamma _{i} \beta _{i}^{\beta _{i}} \Biggr) \bigl(r(b)-r(a) \bigr)^{\alpha -1} \int _{r(a)}^{r(t)}g_{1} (s) \\ &\quad {} \times \biggl[ f(s)+ \int _{r(a)}^{s} \bigl(2^{2-\alpha }r^{\alpha -1}(s)-v^{ \alpha -1} \bigr)g_{2}(v)\,\mathrm{d}v \biggr]\,\mathrm{d}s, \end{aligned} $$

and

$$ B=1- \frac{2\exp [ \frac{c(b)K(b)-2c(a)K(b)}{2} ]}{[\exp (c(b)-2c (a))-4K(b)]^{\frac{1}{2}} +1}. $$

Proof

Note that

$$ \sum^{i=n}_{i=1} \gamma _{i}t^{\beta _{i}} \leq \sum ^{i=n}_{ i=1}\gamma _{i}(\beta _{i} +t)^{\beta _{i}}= \sum^{i=n}_{i=1} \gamma _{i}\beta _{i}^{\beta _{i}} \biggl(1+ \frac{t}{\beta _{ i}} \biggr)^{\beta _{i}} \leq e^{t}\sum ^{ i=n}_{i=1} \gamma _{i}\beta _{i}^{\beta _{i}}. $$
(3.37)

We take \(h(t)=e^{t}\). Then \(Q(t)=1-e^{-t}\), \(Q^{-1} (t)=\log ( \frac{1}{1-t})\), \(S(t)=\exp (-t)-\exp (-2t-c(b) +2c(a))\), \(S^{-1}(t)= \log [\exp (\frac{-c(b)+ 2c(a)}{2} ) (\exp (c(b)-2c(a))-4t )^{\frac{1}{2}}+1 ]-\log (2t)\), and \(Q [S^{-1}(t) ] =1- \frac{2\exp [\frac{c(b)t-2c(a)t}{2} ]}{[\exp (c(b)-2c(a))-4t]^{\frac{1}{2}}+1}\). By applying Theorem 3.3 we deduce the corollary. □

Corollary 3.3

If \(F(y_{1}, y_{2}, y_{3}, y_{4})\leq A(y_{1}, y_{2}) (y_{3}^{p}+y_{4} )\) and \(C(u, v)\leq c(u, v)(u-u_{0}) (v-v_{0})\) for \(u\in [u_{0}, u_{T}]\), \(v\in [v_{0}, v_{T}]\), then \(x(u,v)\) defined by (3.32) satisfies

$$ \begin{aligned} x(u, v)&\leq \frac{1}{k} (u-u_{0})^{\alpha -1}(v-v _{0})^{ \alpha -1}(1-p)^{ \frac{1}{1-p}} \biggl\{ k(u- u_{0})^{2-\alpha }(v-v_{0} )^{2- \alpha }c(u, v) \\ &\quad {} -k_{0}kc(u_{0}, v) +k \int _{u_{0}}^{u} \int ^{v} _{v_{0}}A(s, \tau ) \bigl[1 +D(s, \tau ) \bigr]\,\mathrm{d}\tau \,\mathrm{d}s \biggr\} ^{ \frac{1}{1-p}}, \end{aligned} $$
(3.38)

where

$$ D(s, \tau )= \int ^{s}_{u_{ 0}} \int ^{\tau }_{v_{0}} \bigl(2^{2-\alpha }s^{\alpha -1}- w^{\alpha -1} \bigr) \bigl(2 ^{2- \alpha }\tau ^{\alpha -1} -z^{\alpha -1} \bigr)\,\mathrm{d}w\,\mathrm{d}z, $$
(3.39)

and \(k_{0}\) and k are as in Theorem 3.4.

Proof

From the assumptions of the corollary we can deduce that

$$ \begin{aligned} x(u,v)&\leq (u-u_{0}) (v- v_{0})c(u, v) \\ &\quad {} + \int ^{u}_{u_{0}} \int ^{v}_{v_{0}}(u-s)^{ \alpha -1}(v-\tau )^{ \alpha -1}A(s,\tau ) \biggl[x^{p}( s, \tau ) \\ &\quad {} + \int ^{s}_{u_{0}} \int ^{\tau }_{v_{0}}(s-w)^{\alpha -1}( \tau -z)^{\alpha -1}x^{p}(w,z) \,\mathrm{d}w\,\mathrm{d} z \biggr] \,\mathrm{d}s \,\mathrm{d}\tau . \end{aligned} $$
(3.40)

Applying Theorem 3.4 to \(h(t)=t^{p}\) completes the proof. □

Hermite–Hadamard-type fractional integral inequalities

In this section, we present some Hermite–Hadamard-type fractional integral inequalities by integration, differentiation, and convex functions.

Lemma 4.1

Let \(c, \alpha \in (0, 1)\), and let \(f\in C^{3}([a, b])\). Then

$$ \begin{aligned} _{\mathrm{RL}}{ \mathrm{D}}^{-\alpha } _{a, t}f(t)\big|_{t=b}&= \frac{(b-a)^{\alpha }}{\Gamma (\alpha +3)}T_{ f, c}(a, b)+ \frac{(b-a) ^{\alpha +3}}{\Gamma ( \alpha +3)} \int _{0}^{ 1}Q(t)f^{(3)} \bigl(at+(1- t)b \bigr)\,\mathrm{d}t. \end{aligned} $$
(4.1)

Furthermore, \(Q(t)\) and \(T_{f, c}(a, b)\) can be expressed as follows:

$$ Q(t)=\textstyle\begin{cases} \begin{aligned} Q_{1}(t)=t^{\alpha + 2}-c^{\alpha }t^{2}, & t\in [0,c], \\ Q_{2}(t)=t^{\alpha +2} +b_{2}t^{2}+b_{1}t+b_{0}, & t\in (c,1], \end{aligned} \end{cases} $$
(4.2)

and

$$ T_{f, c}(a, b)=\alpha \frac{2c^{\alpha +1}- 1}{c-1}f(a)+2c^{ \alpha }f(b)+2\frac{ 1-c^{\alpha }}{(c-1) ^{2}}f \bigl(ac+(1-c)b \bigr). $$
(4.3)

Here \(b_{0}=\frac{\alpha c}{2(c-1)}\), \(b_{1}=\frac{\alpha }{1-c}\), \(b_{2}=- \frac{\alpha (c-2)}{2(c-1)}-1\), and \(2(c^{\alpha +1}-1)=(\alpha +2)(c-1)\).

Proof

We represent A as

$$ \begin{aligned} A&= \int _{0}^{1}Q(t) f^{(3)} \bigl(at+(1-t)b \bigr) \,\mathrm{d}t \\ &= \int _{0}^{c}Q_{1}(t) f^{(3)} \bigl(at+(1-t)b \bigr) \,\mathrm{d}t+ \int _{c} ^{1}Q_{2}(t)f^{(3)} \bigl(at+(1-t)b \bigr)\,\mathrm{d}t \\ &=A_{1}+A_{2}. \end{aligned} $$
(4.4)

First, we estimate \(A_{1}\). It is clear that

$$ \begin{aligned} A_{1}&= \int _{0}^{c}Q_{1}(t)f^{(3)} \bigl(at+(1-t) b \bigr) \,\mathrm{d}t \\ &=\frac{1}{a-b} \int _{0} ^{c}Q_{1}(t) \bigl[f^{(2) } \bigl(at+(1-t)b \bigr) \bigr]' \, \mathrm{d}t. \end{aligned} $$
(4.5)

Using integration by parts and the facts \(Q_{1}(0)=Q_{1}(c)=0\), \(Q'_{1}(0)=0\), and \(Q'_{1}(c)=\alpha c^{\alpha +1}\), we have

$$ \begin{aligned} A_{1}&=\frac{1}{b-a} \int _{0}^{c}Q'_{1}(t) f^{(2)} \bigl(at+(1-t)b \bigr) \,\mathrm{d}t \\ &=-\frac{1}{(b-a)^{2}} \int _{0}^{c}Q'_{1}(t) \,\mathrm{d} \bigl[f' \bigl(at +(1-t)b \bigr) \bigr] \\ &=-\frac{\alpha c^{\alpha +1}}{(b-a)^{2}} f' \bigl(ac+(1-c)b \bigr) \\ &\quad {} -\frac{1}{(b-a) ^{3}}Q''_{1}(t)f \bigl(at+ (1-t)b \bigr)\big|_{0}^{c} \\ &\quad {} +\frac{\alpha (\alpha +1)(\alpha +2)}{(b-a)^{3}} \int _{0}^{c}t^{ \alpha -1} f \bigl(at+(1-t)b \bigr)\,\mathrm{d}t. \end{aligned} $$
(4.6)

In a similar manner, we find that

$$ \begin{aligned} A_{2}&=\frac{\alpha c^{ \alpha +1}}{(b-a)^{2}}f' \bigl(ac+(1-c)b \bigr) \\ &\quad {} -\frac{1}{(b-a)^{ 3}}Q''_{2}(t)f \bigl(at+(1-t) b \bigr)\big|_{c}^{1} \\ &\quad {} +\frac{\alpha ( \alpha +1)(\alpha +2)}{(b-a)^{3}} \int _{c}^{1}t^{\alpha -1}f \bigl(at+ (1-t)b \bigr)\,\mathrm{d}t. \end{aligned} $$
(4.7)

Thus A can be written as

$$ \begin{aligned} A&=A_{1}+A_{2} \\ &=-\frac{1}{(b-a)^{3}} \bigl[-Q''_{1}(0)f(b) +Q''_{2}(1)f(a) \bigr] \\ &\quad {} -\frac{1}{(b-a)^{3}} \bigl[Q''_{1}(c)- Q''_{2}(c) \bigr]f \bigl(ac+ (1-c)b \bigr) \\ &\quad {} +\frac{\alpha ( \alpha +1)(\alpha +2)}{ (b-a)^{\alpha +3}} \int _{a}^{b}(b-u)^{\alpha -1}f(u) \, \mathrm{d}u, \end{aligned} $$
(4.8)

which yields the desired result

$$ {}_{\mathrm{RL}}{\mathrm{D}}^{-\alpha } _{a, t}f(t)\big|_{t=b}= \frac{(b-a)^{\alpha }}{\Gamma (\alpha +3)}T_{ f, c}(a, b)+ \frac{(b-a) ^{\alpha +3}}{\Gamma ( \alpha +3)} \int _{0}^{ 1}Q(t)f^{(3)} \bigl(at+(1- t)b \bigr)\,\mathrm{d}t $$
(4.9)

with

$$ \begin{aligned} T_{f, c}(a, b)&=Q''_{2} (1)f(a)-Q''_{1}(0)f(b )+ \bigl[Q''_{1}(c)-Q''_{2} (c) \bigr]f \bigl(ac+(1-c)b \bigr) \\ &=\alpha \frac{2c^{ \alpha +1}-1}{c-1}f(a) +2c^{\alpha }f(b)+2 \frac{1-c^{\alpha }}{(c-1)^{2}}f \bigl(ac+(1- c)b \bigr). \end{aligned} $$
(4.10)

The proof is completed. □

Corollary 4.1

Let \(c, \alpha \in (0, 1)\), and let \(f\in C^{4}([a, b])\). Under the assumptions of Lemma 4.1for \(1-\alpha \), we have

$$ {}_{C}{ \mathrm{D}}^{\alpha }_{a, t}f(t)\big|_{t=b} = \frac{(b-a)^{1-\alpha } }{\Gamma (4-\alpha )}T_{ f', c}(a, b)+ \frac{(b-a) ^{4-\alpha }}{\Gamma (4- \alpha )} \int _{0}^{1}Q (t)f^{(4)} \bigl(at+(1-t)b \bigr) \,\mathrm{d}t. $$
(4.11)

Theorem 4.1

Let \(f\in C^{3}([a, b])\). Denote by d a division of the interval \([a, b]\), i.e., \(d: a=t_{0} < t_{1}<\cdots <t_{n-1}<t_{n}=b\), \(h=t_{i+1} -t_{i}\), \(i=0,\dots ,n-1\). Then for \(c, \alpha \in (0, 1)\),

$$ \begin{aligned} & {}_{\mathrm{RL}}{\mathrm{D}}^{- \alpha }_{a, t}f(t)\big|_{ t=t_{n}} \\ &\quad =\frac{h^{\alpha }}{ \Gamma (\alpha +3)} \Biggl[ \sum_{i=0}^{n-2}T_{F_{i}, c}(t_{i}, t_{i+1}) + T_{f, c}(t_{n-1}, t_{n}) \Biggr]+ \frac{h^{\alpha +3}}{ \Gamma (\alpha +3)} \\ &\qquad {} \times \int _{0}^{1} Q(t) \Biggl[\sum _{i=0}^{n- 2}F_{i}^{(3)} \bigl(t_{i} t+(1- t) t_{i+1} \bigr)+f^{(3)} \bigl(t_{n- 1} t+(1-t) t_{n} \bigr) \Biggr] \,\mathrm{d}t, \end{aligned} $$
(4.12)

where \(F_{i}(u)= (\frac{t_{i+1}-u}{t_{n}-u} )^{1-\alpha }f(u)\), \(i=0,\dots ,n-2\).

Proof

From Lemma 4.1 we have

$$ \begin{aligned} & {}_{\mathrm{RL}}{\mathrm{D}}^{-\alpha } _{a, t}f(t)\big|_{t=t_{n}} \\ &\quad =\frac{1}{\Gamma (\alpha )}\sum_{i=0}^{n-1} \int _{t _{i}}^{t_{i+1}}(t_{n}-u)^{\alpha -1}f(u) \,\mathrm{d}u \\ &\quad =\frac{1}{\Gamma (\alpha )} \sum_{i=0}^{n-2} \int _{t_{i }}^{t_{i+1}}(t_{n}-u)^{ \alpha -1}f(u) \,\mathrm{d}u +\frac{1}{\Gamma (\alpha )} \int _{t_{n-1}}^{t_{n}}(t_{n}-u)^{\alpha -1}f(u) \,\mathrm{d}u \\ &\quad =\frac{1}{\Gamma (\alpha )} \sum_{i=0}^{n-2} \int _{t_{i} }^{t_{i+1}}(t_{i+1}-u)^{ \alpha -1} \biggl(\frac{t_{i +1}-u}{t_{n}-u} \biggr)^{1 -\alpha }f(u)\,\mathrm{d}u \\ &\qquad {} +\frac{1}{\Gamma ( \alpha )} \int _{t_{n-1}}^{ t_{n}}(t_{n}-u)^{\alpha - 1}f(u) \,\mathrm{d}u \\ &\quad =\frac{1}{\Gamma (\alpha ) }\sum_{i=0}^{n-2} \int _{t _{i}}^{t_{i+1}}(t_{i+1}- u)^{\alpha -1}F_{i}(u) \,\mathrm{d}u+\frac{1}{ \Gamma (\alpha )} \int _{t_{n-1}}^{t_{n}}(t_{n}-u) ^{\alpha -1}f(u) \,\mathrm{d}u \\ &\quad =\sum_{i=0}^{n-2} \frac{(t_{i+1}-t_{i})^{\alpha }}{\Gamma (\alpha +3)}T_{ F_{i}, c}(t_{i}, t_{i+1}) +\sum_{i=0}^{n-2} \frac{ (t_{i+1}-t_{i})^{\alpha +3}}{\Gamma (\alpha +3)} \\ &\qquad {} \times \int _{0}^{1} Q(t)F_{i}^{(3)} \bigl(t_{i} t+ (1-t) t_{i+1} \bigr) \,\mathrm{d} t+ \frac{(t_{n}-t_{n-1}) ^{\alpha }}{\Gamma (\alpha +3)}T_{f, c}(t_{n-1},t_{n}) \\ &\qquad {} +\frac{(t_{n}-t_{n -1})^{\alpha +3}}{\Gamma ( \alpha +3)} \int _{0}^{1}Q(t) f^{(3)} \bigl(t_{n-1} t+(1-t) t_{n} \bigr) \,\mathrm{d}t \\ &\quad =\frac{h^{\alpha }}{\Gamma (\alpha +3)}\sum_{i=0}^{n- 2}T_{F_{i}, c}(t_{i}, t_{i+ 1})+\frac{h^{\alpha }}{ \Gamma (\alpha +3)}T_{f, c} (t_{n-1}, t_{n}) \\ &\qquad {} +\frac{h^{\alpha +3}}{ \Gamma (\alpha +3)} \int _{0} ^{1}Q(t) \Biggl[\sum _{ i=0}^{n-2}F_{i}^{(3)} \bigl(t_{i} t+(1-t) t_{i+1} \bigr) \Biggr] \,\mathrm{d}t \\ &\qquad {} +\frac{h^{ \alpha +3}}{\Gamma (\alpha +3)} \int _{0}^{1}Q(t)f^{ (3)} \bigl(t_{n-1} t+(1-t) t_{n} \bigr)\,\mathrm{d}t \\ &\quad =\frac{h^{\alpha }}{\Gamma (\alpha +3)} \Biggl[\sum_{ i=0}^{n-2}T_{F_{i}, c}(t _{i}, t_{i+1})+T_{f, c}(t _{n-1}, t_{n}) \Biggr]+ \frac{h^{\alpha +3}}{ \Gamma (\alpha +3)} \\ &\qquad {} \times \int _{0}^{ 1}Q(t) \Biggl[\sum _{i=0} ^{n-2}F_{i}^{(3)} \bigl(t_{i} t+(1-t) t_{i+1} \bigr)+f^{(3)} \bigl(t_{n-1} t+(1-t) t_{n} \bigr) \Biggr] \,\mathrm{d}t . \end{aligned} $$

This proves the theorem. □

Theorem 4.2

Let \(f\in C^{4}([a, b])\) and suppose that \(|f^{(3)} |\) is an \((s, m)\)-convex function on \([a, b]\). Under the assumptions of Lemma 4.1, we have the following inequality for \(c, \alpha \in (0, 1)\):

$$\begin{aligned} & \biggl\lvert _{\mathrm{RL}}{ \mathrm{D}}^{-\alpha }_{a, t}f(t)\big|_{t=b}- \frac{(b- a)^{\alpha }}{\Gamma ( \alpha +3)}T_{f, c}(a, b) \biggr\rvert \\ &\quad \leq \frac{(b-a)^{\alpha +3}}{\Gamma (\alpha +3)} \biggl\{ (1+b_{1}) \biggl[ \frac{\alpha c^{\alpha + s+3}-\alpha (1-c)^{\alpha +s+3}}{(s+3)(\alpha +s+3)} +3 \biggr] \bigl( \bigl\vert f^{ (3)}(a) \bigr\vert + \bigl\vert f^{( 3)}(b) \bigr\vert \bigr) \\ &\qquad {} +2^{2-s}m(1+b_{1}) \biggl(\frac{c^{\alpha +s+ 3}}{\alpha +s+3}+ \frac{c^{s+1}}{s+1} \biggr) \biggl( \biggl\vert f^{(3)} \biggl( \frac{a}{m} \biggr) \biggr\vert + \biggl\lvert f^{(3)} \biggl( \frac{b}{m} \biggr) \biggr\rvert \biggr) \\ &\qquad {} -\mu \biggl[ \biggl(a-\frac{b}{m} \biggr) ^{2}+ \biggl( b- \frac{a}{m} \biggr)^{2} \biggr] \int _{0} ^{1}t^{3} \bigl(c^{\alpha }-t^{ \alpha } \bigr) (1-t)\, \mathrm{d}t \biggr\} . \end{aligned}$$
(4.13)

Proof

Equation (4.1) yields the inequality

$$ \begin{aligned} & \biggl\lvert {}_{\mathrm{RL}}{ \mathrm{D}}^{-\alpha }_{a, t}f(t)\big|_{t=b}- \frac{(b- a)^{\alpha }}{\Gamma ( \alpha +3)}T_{f, c}(a, b) \biggr\rvert \\ &\quad \leq \frac{(b-a)^{\alpha +3}}{\Gamma (\alpha +3)} \int _{0}^{1} \bigl\vert Q(t) \bigr\vert \bigl\vert f^{(3)} \bigl(at +(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \\ &\quad =\frac{(b-a)^{\alpha +3} }{\Gamma (\alpha +3)} \biggl[ \int _{0}^{c} \bigl\vert Q_{ 1}(t) \bigr\vert \bigl\vert f^{(3)} \bigl(at+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \\ &\qquad {} + \int _{c}^{ 1} \bigl\vert Q_{2}(t) \bigr\vert \bigl\vert f^{(3)} \bigl(at+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \biggr] \\ &\quad =\frac{(b-a)^{\alpha +3}}{\Gamma (\alpha +3)}(I_{1}+ I_{2}). \end{aligned} $$
(4.14)

Since \(| f^{(3)} |\) is \((s,m)\)-convex, \(I_{1}\) can be calculated as

$$ \begin{aligned} I_{1}&= \int _{0}^{c} \bigl\vert Q_{1}(t) \bigr\vert \bigl\vert f^{(3)} \bigl(at+(1-t)b \bigr) \bigr\vert \,\mathrm{d} t \\ &= \int _{0}^{c} t^{2} \bigl(c^{ \alpha }-t^{\alpha } \bigr) \bigl\vert f^{(3)} \bigl(at+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \\ &\leq \int _{0}^{c} t^{2} \bigl(c^{\alpha }-t^{\alpha } \bigr) \biggl\{ t^{s} \bigl\vert f^{(3)} (a) \bigr\vert +m(1-t)^{s} \biggl\lvert f^{(3)} \biggl( \frac{b}{m} \biggr) \biggr\rvert \\ &\quad {} -\mu t(1-t) \biggl(a-\frac{b}{m} \biggr) ^{2} \biggr\} \,\mathrm{d}t \\ &= \bigl\vert f^{(3)}(a) \bigr\vert \int _{0}^{c}t^{s+2} \bigl(c^{ \alpha }-t^{\alpha } \bigr) \,\mathrm{d}t+m \biggl\lvert f^{(3)} \biggl(\frac{b}{m} \biggr) \biggr\rvert \int _{0}^{c}t^{2} \bigl(c^{\alpha } -t^{\alpha } \bigr) (1-t)^{s} \, \mathrm{d}t \\ &\quad {} -\mu \biggl(a-\frac{b}{m} \biggr)^{2} \int _{0}^{c}t^{3} \bigl(c^{ \alpha } -t^{\alpha } \bigr) (1-t)\,\mathrm{d}t \\ &= \frac{\alpha c^{s+ \alpha +3} \vert f^{(3)} (a) \vert }{(s+3)(s+\alpha +3)}+m \biggl\lvert f^{ (3)} \biggl( \frac{b}{m} \biggr) \biggr\rvert \int _{0}^{c}t^{2} \bigl(c^{\alpha } -t^{ \alpha } \bigr) (1-t)^{s} \, \mathrm{d}t \\ &\quad {} -\mu \biggl( a-\frac{b}{m} \biggr)^{2} \int _{0 }^{c}t^{3} \bigl(c^{ \alpha }-t^{\alpha } \bigr) (1-t)\,\mathrm{d} t. \end{aligned} $$

Furthermore, we have

$$ \begin{aligned} &\int _{0}^{c}t^{2} \bigl(c^{\alpha }-t^{\alpha } \bigr) (1-t)^{s} \, \mathrm{d}t \\ &\quad \leq 2^{1-s} \int _{0}^{c} t^{2} \bigl(1-t^{\alpha } \bigr) \bigl(1-t^{s} \bigr)\,\mathrm{d}t \\ &\quad =2^{1-s} \biggl( \frac{c ^{\alpha +s+3}}{\alpha +s +3}-\frac{c^{s+3}}{s+3} + \frac{c^{s+2}}{s+2} \biggr). \end{aligned} $$

Hence it is easily shown that

$$ \begin{aligned} I_{1}&\leq \frac{\alpha c^{s+\alpha +3} \vert f ^{(3)}(a) \vert }{(s+3)( s+\alpha +3)}+2^{1-s}m \biggl(\frac{c^{\alpha + s+3}}{\alpha +s+3}- \frac{c^{s+3}}{s+3}+\frac{c^{s+2}}{s+2} \biggr) \biggl\lvert f^{(3)} \biggl( \frac{b}{m} \biggr) \biggr\rvert \\ &\quad {} -\mu \biggl(a-\frac{b}{m} \biggr)^{2} \int _{ 0}^{c}t^{3} \bigl(c^{ \alpha }- t^{\alpha } \bigr) (1-t) \,\mathrm{d}t \\ &\leq \frac{\alpha c^{ s+\alpha +3} \vert f^{ (3)}(a) \vert }{(s+3)(s +\alpha +3)}+2^{1-s}m \biggl(\frac{c^{\alpha + s+3}}{\alpha +s+3}- \frac{c^{s+3}}{s+3}+\frac{c ^{s+1}}{s+1} \biggr) \biggl\lvert f^{(3)} \biggl(\frac{b}{m} \biggr) \biggr\rvert \\ &\quad {} -\mu \biggl(a-\frac{b}{m} \biggr)^{2} \int _{ 0}^{c}t^{3} \bigl(c^{ \alpha }- t^{\alpha } \bigr) (1-t)\,\mathrm{d}t. \end{aligned} $$
(4.15)

Next, using \(c^{\alpha }-t^{\alpha }\leq \alpha (c-t)t^{\alpha -1}\), we have

$$ \begin{aligned} I_{2}&= \int _{c}^{1} \bigl\vert Q_{2}(t) \bigr\vert \bigl\vert f^{(3) } \bigl(at+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \\ &\leq \bigl\vert f^{(3)}(b) \bigr\vert \int _{0}^{1-c}t^{s} \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d} t+m \biggl\vert f^{(3)} \biggl(\frac{a}{m} \biggr) \biggr\vert \int _{0}^{1-c}(1-t) ^{s} \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d}t \\ &\quad {} -\mu \biggl(b-\frac{a}{m} \biggr)^{2} \int _{0}^{1-c} t(1-t) \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d} t. \end{aligned} $$
(4.16)

As for \(|Q_{2}(1-t) |\), we have the inequality

$$ (1-b_{0}) \bigl[m(t)-t+2 \bigr]\leq \bigl\vert Q_{2}(1-t) \bigr\vert \leq (b_{1}+1) \bigl[m(t)+1 \bigr], $$
(4.17)

in which \(m(t)=(1-t)^{\alpha +2}+(1-t)^{2}\), and \((1-t)^{r}+t^{r}\leq 1\) for \(r\geq 1\).

Thus

$$ \begin{aligned} I_{2}&\leq (1+b_{1}) \biggl[ \frac{(1-c)^{ \alpha +s+3}}{\alpha + s+3}+\frac{3(1-c)^{s +1}}{s+1}- \frac{(1-c) ^{s+3}}{s+3} \biggr] \bigl\vert f^{(3)}(b) \bigr\vert \\ &\quad {} +m(1+b_{1}) \biggl(\frac{c^{ \alpha +s+3}}{\alpha + s+3}+\frac{c^{s+3}}{s+3}+ \frac{c^{s+1}}{s+1} \biggr) \biggl\vert f^{(3)} \biggl( \frac{a}{m} \biggr) \biggr\vert \\ &\quad {} -\mu \biggl(b- \frac{a}{m} \biggr)^{2} \int _{0}^{1-c} t(1-t) \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d}t \\ &\leq (1+b_{1}) \biggl[ \frac{-\alpha (1-c)^{ \alpha +s+3}}{(s+3)( \alpha +s+3)}+3 \biggr] \bigl\vert f^{(3)}(b) \bigr\vert \\ &\quad {} +m(1+b_{1}) \biggl(\frac{c^{\alpha +s+3}}{\alpha +s+3}+ \frac{c^{s+3}}{s+3} + \frac{c^{s+1}}{s+1} \biggr) \biggl\vert f^{(3)} \biggl( \frac{a}{m} \biggr) \biggr\vert \\ &\quad {} -\mu \biggl(b- \frac{a}{m} \biggr)^{2} \int _{0}^{1-c}t( 1-t) \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d}t, \end{aligned} $$
(4.18)

and

$$\begin{aligned} &\int _{0}^{1-c}t(1-t) \bigl\vert Q_{2}(1-t) \bigr\vert \,\mathrm{d}t \\ &\quad = \int _{c}^{1}t(1-t) \bigl\vert Q_{2}(t) \bigr\vert \,\mathrm{d}t \\ &\quad \geq \int _{c}^{1}t(1 -t) \bigl(t^{\alpha +2}-b_{ 2}t^{2} \bigr)\,\mathrm{d}t \\ &\quad \geq \int _{c}^{1}t^{ 3}(1-t) \bigl(c^{\alpha }-t ^{\alpha } \bigr)\,\mathrm{d} t. \end{aligned}$$
(4.19)

By (4.15), (4.18), and (4.19) we get the inequality

$$ \begin{aligned} I_{1}+I_{2}& \leq (1+b _{1}) \biggl[ \frac{ \alpha c^{\alpha +s+3} -\alpha (1-c)^{\alpha +s +3}}{(s+3)(\alpha +s+3) }+3 \biggr] \bigl[ \bigl\vert f^{(3)}(a) \bigr\vert + \bigl\vert f^{(3)}(b) \bigr\vert \bigr] \\ &\quad {} +2^{2-s}m(1+b_{1}) \biggl( \frac{c^{ \alpha +s+3}}{\alpha + s+3}+ \frac{c^{s+1}}{s+1} \biggr) \biggl[ \biggl\vert f^{(3)} \biggl( \frac{a}{m} \biggr) \biggr\vert + \biggl\vert f^{(3)} \biggl(\frac{b}{m} \biggr) \biggr\vert \biggr] \\ &\quad {} -\mu \biggl[ \biggl(a-\frac{b}{m} \biggr) ^{2}+ \biggl(b- \frac{a}{m} \biggr)^{2} \biggr] \int _{0}^{1}t^{3} \bigl(c^{ \alpha }-t^{\alpha } \bigr) (1 -t) \,\mathrm{d}t. \end{aligned} $$
(4.20)

Finally, the proof can be fulfilled by (4.14) and (4.20). □

Lemma 4.2

Let \(f\in C([a, b])\). Then for \(c, \alpha \in (0, 1)\),

$$ \begin{aligned} {}_{\mathrm{RL}}{ \mathrm{D}}^{-\alpha } _{a, t}f(t)\big|_{t=b}&= \frac{(b-a)^{\alpha }}{\Gamma (\alpha +1)}f \bigl(c a+(1-c)b \bigr) \\ &\quad {} +\frac{(b-a)^{ \alpha +1}}{\Gamma ( \alpha +1)} \int _{0}^{ 1}p(t)f' \bigl(ta+(1-t)b \bigr) \,\mathrm{d}t, \end{aligned} $$
(4.21)

where

$$ p(t)=\textstyle\begin{cases} t^{\alpha }, & t\in [0, c], \\ t^{\alpha }-1, & t\in (c, 1]. \end{cases} $$
(4.22)

Proof

Let \(J=\int _{0}^{1}p(t)f'(ta+(1-t)b)\,\mathrm{d}t\). We have

$$ \begin{aligned} J&= \int _{0}^{c}t^{\alpha }f' \bigl(ta+(1-t)b \bigr)\,\mathrm{d}t+ \int _{c}^{1} \bigl(t^{ \alpha }-1 \bigr)f' \bigl(ta+(1-t) b \bigr)\,\mathrm{d}t \\ &=\frac{t^{\alpha }}{a-b} f \bigl(ta+(1-t)b \bigr)\big|_{t=0} ^{t=c}- \frac{\alpha }{a- b} \int _{0}^{c}t^{\alpha -1}f \bigl(ta+(1-t)b \bigr)\,\mathrm{d}t \\ &\quad {} +\frac{t^{\alpha }- 1}{a-b}f \bigl(ta+(1-t)b \bigr)| _{t=c}^{t=1}- \frac{ \alpha }{a-b} \int _{c}^{1} t^{\alpha -1}f \bigl(ta+(1-t)b \bigr) \,\mathrm{d}t \\ &=\frac{f(ca+(1-c)b)}{a -b}+\frac{\alpha }{b-a} \int _{0}^{1}t^{\alpha -1 }f \bigl(ta+(1-t)b \bigr) \,\mathrm{d} t \\ &=\frac{f(ca+(1-c)b)}{a -b}+\frac{\alpha }{(b-a) ^{\alpha +1}} \int _{a}^{b}(b-t)^{\alpha -1}f(t) \, \mathrm{d}t. \end{aligned} $$
(4.23)

Multiplying both sides of (4.23) by \(\frac{1}{\Gamma (\alpha )}\), we obtain (4.21). The lemma is proved. □

Theorem 4.3

Let \(c, \alpha \in (0, 1)\), let \(f:[a, b]\mapsto \mathbb{R}\) be a differentiable function on \([a, b]\), and let \(p\geq 1\). If \(\vert f'\vert ^{\frac{p}{p-1}}\) is convex, then

$$ \begin{aligned} & \biggl\lvert _{\mathrm{RL}} {\mathrm{D}}^{-\alpha }_{a, t}f(t)\big|_{t=b}- \frac{( b-a)^{\alpha }}{\Gamma (\alpha +1)}f \bigl(ca+(1-c) b \bigr) \biggr\rvert \\ &\quad \leq \frac{2^{\frac{1}{p}}(b-a)^{\alpha +1}}{\Gamma (\alpha +1)} \bigl( \bigl\vert f'(a) \bigr\vert ^{ \frac{p}{p-1}}+ \bigl\vert f'(b) \bigr\vert ^{\frac{p}{p -1}} \bigr)^{\frac{p- 1}{p}}. \end{aligned} $$
(4.24)

Proof

By (4.21) we have

$$ \begin{aligned} & \biggl\vert {}_{\mathrm{RL}} {\mathrm{D}}^{-\alpha }_{a, t}f(t)\big|_{t=b}- \frac{ (b-a)^{\alpha }}{ \Gamma (\alpha +1)}f \bigl( ca+(1-c)b \bigr) \biggr\rvert \\ &\quad \leq \frac{(b-a)^{ \alpha +1}}{\Gamma ( \alpha +1)} \int _{0} ^{c}t^{\alpha } \bigl\vert f' \bigl(ta+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t+ \int _{c}^{1} \bigl(1-t^{\alpha } \bigr) \bigl\vert f' \bigl(ta+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \\ &\quad \leq \frac{(b-a)^{ \alpha +1}}{\Gamma ( \alpha +1)} \biggl[ \int _{0}^{c}t^{\alpha } \bigl\vert f' \bigl(ta+(1-t)b \bigr) \bigr\vert \,\mathrm{d}t+ \int _{c}^{1} \bigl(1-t^{ \alpha } \bigr) \bigl\vert f' \bigl(ta +(1-t)b \bigr) \bigr\vert \,\mathrm{d}t \biggr]. \end{aligned} $$
(4.25)

Using the Hölder inequality for \(q, p\geq 1\) such that \(\frac{1}{p}+\frac{1}{q}=1\), we can prove that

$$ \begin{aligned} & \biggl\vert {}_{\mathrm{RL}}{ \mathrm{D}}^{-\alpha }_{a, t}f(t) \big\vert _{t=b}- \frac{( b-a)^{\alpha }}{\Gamma ( \alpha +1)}f \bigl(ca+(1-c)b \bigr) \biggr\vert \\ &\quad \leq \frac{(b-a)^{\alpha +1}}{\Gamma (\alpha +1)} \biggl\{ \biggl( \int _{0}^{c}t^{\alpha p} \,\mathrm{d}t \biggr)^{\frac{1}{p}} \biggl[ \int _{0}^{c} \bigl\vert f' \bigl(ta+(1-t)b \bigr) \bigr\vert ^{q} \,\mathrm{d}t \biggr]^{\frac{1}{q}} \\ &\qquad {} + \biggl[ \int _{c}^{1} \bigl(1-t^{ \alpha } \bigr)^{p}\,\mathrm{d}t \biggr]^{\frac{ 1}{p}} \biggl[ \int _{c} ^{1} \bigl\vert f' \bigl(ta+(1-t) b \bigr) \bigr\vert ^{q} \,\mathrm{d}t \biggr]^{\frac{1}{q}} \biggr\} \\ &\quad \leq \frac{(b-a)^{\alpha +1}}{2^{\frac{1}{q}} \Gamma (\alpha +1)} \biggl\{ \biggl( \int _{0}^{c}t ^{\alpha p} \,\mathrm{d} t \biggr)^{\frac{1}{p}} \bigl[c^{2} \bigl\vert f'(a) \bigr\vert ^{q}+ \bigl(1-(1-c)^{ 2} \bigr) \bigl\vert f'(b) \bigr\vert ^{ q} \bigr]^{\frac{1}{q}} \\ &\qquad {} + \biggl[ \int _{c}^{1} \bigl(1-t^{\alpha } \bigr)^{p}\,\mathrm{d}t \biggr] ^{\frac{1}{p}} \bigl[ \bigl(1 -c^{2} \bigr) \bigl\vert f'(a) \bigr\vert ^{q}+(1-c)^{2} \bigl\vert f' (b) \bigr\vert ^{q} \bigr]^{ \frac{1}{q}} \biggr\} \\ &\quad \leq \frac{(b-a)^{\alpha +1}}{2^{\frac{1}{q}} \Gamma (\alpha +1)} \biggl\{ \biggl( \int _{0} ^{c}t^{\alpha p} \, \mathrm{d}t \biggr)^{\frac{1}{p}}+ \biggl[ \int _{c}^{1} \bigl(1-t^{\alpha } \bigr)^{p} \,\mathrm{d}t \biggr]^{ \frac{1}{p}} \biggr\} \\ &\qquad {} \times \bigl\{ \bigl[c^{2} \bigl\vert f'(a) \bigr\vert ^{q}+ \bigl(1-(1-c)^{2} \bigr) \bigl\vert f'(b) \bigr\vert ^{q} \bigr]^{\frac{1}{q}} \\ &\qquad {} + \bigl[ \bigl(1 -c^{2} \bigr) \bigl\vert f'(a) \bigr\vert ^{q}+(1-c)^{2} \bigl\vert f'(b) \bigr\vert ^{q} \bigr] ^{\frac{1}{q}} \bigr\} \\ &\quad \leq \frac{(b-a)^{\alpha +1}}{2^{\frac{1}{q}} \Gamma (\alpha +1)}2^{1 -1/p} \biggl( \int _{0}^{c} t^{\alpha p} \,\mathrm{d} t+ \int _{c}^{1} \bigl(1-t^{ \alpha } \bigr)^{p}\,\mathrm{d} t \biggr)^{ \frac{1}{p}} 2^{1-1/q} \bigl( \bigl\vert f'(a) \bigr\vert ^{q}+ \bigl\vert f'(b) \bigr\vert ^{q} \bigr) ^{\frac{1}{q}} \\ &\quad =\frac{2^{\frac{1}{ p}}(b-a)^{\alpha +1}}{\Gamma (\alpha +1)} \biggl( \int _{0}^{c}t ^{\alpha p}\,\mathrm{d} t+ \int _{c}^{1} \bigl(1-t^{ \alpha } \bigr)^{p}\,\mathrm{d}t \biggr)^{\frac{1}{p}} \bigl( \bigl\vert f'(a) \bigr\vert ^{q}+ \bigl\vert f'(b) \bigr\vert ^{q} \bigr)^{ \frac{1}{q}} \\ &\quad \leq \frac{2^{\frac{ 1}{p}}(b-a)^{\alpha +1 }}{\Gamma (\alpha +1)} \bigl( \bigl\vert f'(a) \bigr\vert ^{q}+ \bigl\vert f'(b) \bigr\vert ^{q} \bigr)^{ \frac{1}{q}} , \end{aligned} $$
(4.26)

where we used the inequalities \(A^{r}+B^{r}\leq 2^{1-r}(A+B)^{r}\) for \(A, B\geq 0\) and \(0\leq r \leq 1\), and \((1-t^{\alpha })^{p}+t^{\alpha p}\leq 1\) for \(0\leq t \leq 1\). The proof is completed. □

Proposition 4.1

Let \(p, \beta \geq 0, c\), \(\alpha \in (0,1)\), and \(0\leq a\leq b\). Then

$$ \bigl\vert \mathrm{B}(\alpha , \beta ) (b-a)^{\beta }- \bigl(ca +(1-c)b \bigr)^{ \beta } \bigr\vert \leq \beta ^{\frac{p}{p- 1}}2^{\frac{1}{p}}(b-a) \bigl(a^{\frac{p(\beta - 1)}{p-1}}+b^{\frac{p( \beta -1)}{p-1}} \bigr), $$
(4.27)

where \(\mathrm{B} (\alpha ,\beta )=\int _{0} ^{1}u^{\alpha -1}(1-u)^{ \beta -1}\,\mathrm{d}u\).

Proof

The proposition comes directly by applying Theorem 4.3 to \(f(t)=t^{\beta }\). □

Proposition 4.2

Let \(\beta \geq 3\) and \(c, \alpha \in (0,1)\) with \((\alpha +2)(c-1)=2(c^{\alpha +1}-1)\). Then

$$ \biggl\vert \mathrm{B} (\alpha , \beta ) (b-a)^{\beta }- \frac{T_{t^{\beta }, c}(a,b)}{ \alpha (\alpha +1)( \alpha +2)} \biggr\vert \leq \frac{\beta ( \beta -1)(\beta -2)}{\alpha (\alpha +1) (\alpha +2)}(b-a)^{3} \int _{0}^{1} \bigl\vert Q(u) \bigr\vert \,\mathrm{d}u. $$
(4.28)

Proof

Applying Lemma 4.1 to \(f(t)=t^{\beta }\) and \(\beta \geq 3\), we arrive at (4.28). □

Conclusion

In this paper, we established some new Volterra–Fredholm and Hermite–Hadamard-type fractional integral inequalities. They extend some known inequalities and provide a handy tool for deriving bounds of solutions to fractional differential equations and fractional integral equations. In the meantime, we obtain new fractional integral inequalities for convex functions and show their applications. Finally, we present some estimates of the Riemann–Liouville fractional integral of functions whose absolute value is convex and the derivative is raised to a positive real power.

Availability of data and materials

Not applicable.

References

  1. Li, C.P., Li, Z.Q., Wang, Z.: Mathematical analysis and the local discontinuous Galerkin method for Caputo–Hadamard fractional partial differential equation. J. Sci. Comput. 85(2), 41 (2020)

    MathSciNet  Article  Google Scholar 

  2. Li, C.P., Li, Z.Q.: Stability and logarithmic decay of the solution to Hadamard-type fractional differential equation. J. Nonlinear Sci. 31(2), 31 (2021)

    MathSciNet  Article  Google Scholar 

  3. Li, C.P., Li, Z.Q.: The blow-up and global existence of solution to Caputo–Hadamard fractional partial differential equation with fractional Laplacian. J. Nonlinear Sci. 31(5), 80 (2021)

    MathSciNet  Article  Google Scholar 

  4. Fan, E.Y., Li, C.P., Li, Z.Q.: Numerical approaches to Caputo–Hadamard fractional derivatives with application to long-term integration of fractional systems. Commun. Nonlinear Sci. Numer. Simul. 106, 106096 (2022)

    MathSciNet  Article  Google Scholar 

  5. Abdeldaim, A., El-Deeb, A.A.: On generalized of certain retarded nonlinear integral inequalities and its applications in retarded integro-differential equations. Appl. Math. Comput. 256, 375–380 (2015)

    MathSciNet  MATH  Google Scholar 

  6. Boudeliou, A.: On certain new nonlinear retarded integral inequalities in two independent variables and applications. Appl. Math. Comput. 335, 103–111 (2018)

    MathSciNet  MATH  Google Scholar 

  7. El-Deeb, A.A., Ahmed, R.G.: On some generalizations of certain nonlinear retarded integral inequalities for Volterra–Fredholm integral equations and their applications in delay differential equations. J. Egypt. Math. Soc. 25, 279–285 (2017)

    MathSciNet  Article  Google Scholar 

  8. Gu, J., Meng, F.W.: Some new nonlinear Volterra–Fredholm type dynamic integral inequalities on time scales. Appl. Math. Comput. 245, 235–242 (2014)

    MathSciNet  MATH  Google Scholar 

  9. Liu, H.D.: A class of retarded Volterra–Fredholm type integral inequalities on time scales and their applications. J. Inequal. Appl. 2017(1), 293 (2017)

    MathSciNet  Article  Google Scholar 

  10. Xu, R., Ma, X.T.: Some new retarded nonlinear Volterra–Fredholm type integral inequalities with maxima in two variables and their applications. J. Inequal. Appl. 2017(1), 187 (2017)

    MathSciNet  Article  Google Scholar 

  11. Bracamonte, M., Giménez, J., Vivas-Cortez, M.: Hermite–Hadamard–Fejér type inequalities for \((s, m)\)-strongly convex functions with module c, in the second sense. Appl. Math. Inf. Sci. 10(6), 2015–2053 (2016)

    MathSciNet  Article  Google Scholar 

  12. Alabdali, O., Guessab, A., Schmeisser, G.: Characterizations of uniform convexity for differentiable functions. Appl. Anal. Discrete Math. 13(3), 721–732 (2019)

    MathSciNet  Article  Google Scholar 

  13. Guessab, A., Moncayo, M., Schmeisser, G.: A class of nonlinear four-point subdivision schemes. Adv. Comput. Math. 37(2), 151–190 (2012)

    MathSciNet  Article  Google Scholar 

  14. Guessab, A.: Direct and converse results for generalized multivariate Jensen-type inequalities. J. Nonlinear Convex Anal. 13(4), 777–797 (2012)

    MathSciNet  MATH  Google Scholar 

  15. Guessab, A.: Sharp Approximations Based on Delaunay Triangulations and Voronoi Diagrams: Textbook, Novosibirsk State University. NSU Publishing and Printing Center, Novosibirsk (2022)

    Google Scholar 

  16. Guessab, A.: Generalized barycentric coordinates and approximations of convex functions on arbitrary convex polytopes. Comput. Math. Appl. 66(6), 1120–1136 (2013)

    MathSciNet  Article  Google Scholar 

  17. Deng, J.H., Wang, J.R.: Fractional Hermite–Hadamard inequalities for \((\alpha , m)\)-logarithmically convex functions. J. Inequal. Appl. 2013, 364 (2013)

    MathSciNet  Article  Google Scholar 

  18. Hudzik, H., Maligranda, L.: Some remarks on s-convex functions. Aequ. Math. 48, 100–111 (1994)

    MathSciNet  Article  Google Scholar 

  19. Iscan, I.: Hermite–Hadamard’s inequalities for prequasiinvex functions via fractional integrals. Konuralp J. Math. 2(2), 76–84 (2014)

    MATH  Google Scholar 

  20. Set, E., Özdemir, M.E., Dragomir, S.S.: On Hadamard-type inequalities involving several kinds of convexity. J. Inequal. Appl. 2010, Article ID 286845 (2010)

    MathSciNet  Article  Google Scholar 

  21. Li, C.P., Cai, M.: Theory and Numerical Approximations of Fractional Integrals and Derivatives. SIAM, Philadelphia (2019)

    Book  Google Scholar 

  22. Li, C.P., Zeng, F.H.: Numerical Methods for Fractional Calculus. Chapman & Hall/CRC, Boca Raton (2015)

    Book  Google Scholar 

  23. Podlubny, I.: Fractional Differential Equations. Academic Press, San Diego (1999)

    MATH  Google Scholar 

  24. Alomari, M., Darus, M.: The Hadamard’s inequality for s-convex function. Int. J. Math. Anal. 2(13), 639–646 (2008)

    MathSciNet  MATH  Google Scholar 

  25. Hadamard, J.: Étude sur les propriétés des fonctions entières et en particulier d’une fonction considérée par Riemann. J. Math. Pures Appl. 58, 171–215 (1893)

    MATH  Google Scholar 

Download references

Acknowledgements

The authors would like to express their sincere thanks to the editor and the anonymous reviewers for their helpful comments and suggestions.

Funding

Not applicable.

Author information

Affiliations

Authors

Contributions

MDB made the major analysis and the original draft preparation. JT contributed significantly in writing this paper by analyzing the results, reviewing and editing. Both authors read and approved the final manuscript.

Corresponding author

Correspondence to Jianhua Tang.

Ethics declarations

Competing interests

The authors declare that they have no competing interests.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Doubbi Bounoua, M., Tang, J. On some Volterra–Fredholm and Hermite–Hadamard-type fractional integral inequalities. J Inequal Appl 2022, 36 (2022). https://doi.org/10.1186/s13660-022-02772-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13660-022-02772-6

Keywords

  • Volterra–Fredholm inequalities
  • Hermite–Hadamard inequalities
  • Fractional integral inequalities