In this work, the existence of solutions for nonlinear hybrid fractional integro-differential equations involving generalized proportional fractional (GPF) derivative of Caputo-Liouville-type and multi-term of GPF integrals of Reimann-Liouville type with Dirichlet boundary conditions is investigated. The analysis is accomplished with the aid of the Dhage's fixed point theorem with three operators and the lower regularized incomplete gamma function. Further, the uniqueness of solutions and their Ulam-Hyers-Rassias stability to a special case of the suggested hybrid problem are discussed. For the sake of corroborating the obtained results, an illustrative example is presented.
Citation: Zaid Laadjal, Fahd Jarad. Existence, uniqueness and stability of solutions for generalized proportional fractional hybrid integro-differential equations with Dirichlet boundary conditions[J]. AIMS Mathematics, 2023, 8(1): 1172-1194. doi: 10.3934/math.2023059
[1] | Qun Dai, Shidong Liu . Stability of the mixed Caputo fractional integro-differential equation by means of weighted space method. AIMS Mathematics, 2022, 7(2): 2498-2511. doi: 10.3934/math.2022140 |
[2] | Murugesan Manigandan, R. Meganathan, R. Sathiya Shanthi, Mohamed Rhaima . Existence and analysis of Hilfer-Hadamard fractional differential equations in RLC circuit models. AIMS Mathematics, 2024, 9(10): 28741-28764. doi: 10.3934/math.20241394 |
[3] | Weerawat Sudsutad, Chatthai Thaiprayoon, Sotiris K. Ntouyas . Existence and stability results for $ \psi $-Hilfer fractional integro-differential equation with mixed nonlocal boundary conditions. AIMS Mathematics, 2021, 6(4): 4119-4141. doi: 10.3934/math.2021244 |
[4] | Naimi Abdellouahab, Keltum Bouhali, Loay Alkhalifa, Khaled Zennir . Existence and stability analysis of a problem of the Caputo fractional derivative with mixed conditions. AIMS Mathematics, 2025, 10(3): 6805-6826. doi: 10.3934/math.2025312 |
[5] | Ugyen Samdrup Tshering, Ekkarath Thailert, Sotiris K. Ntouyas . Existence and stability results for a coupled system of Hilfer-Hadamard sequential fractional differential equations with multi-point fractional integral boundary conditions. AIMS Mathematics, 2024, 9(9): 25849-25878. doi: 10.3934/math.20241263 |
[6] | Songkran Pleumpreedaporn, Chanidaporn Pleumpreedaporn, Weerawat Sudsutad, Jutarat Kongson, Chatthai Thaiprayoon, Jehad Alzabut . On a novel impulsive boundary value pantograph problem under Caputo proportional fractional derivative operator with respect to another function. AIMS Mathematics, 2022, 7(5): 7817-7846. doi: 10.3934/math.2022438 |
[7] | Bounmy Khaminsou, Weerawat Sudsutad, Jutarat Kongson, Somsiri Nontasawatsri, Adirek Vajrapatkul, Chatthai Thaiprayoon . Investigation of Caputo proportional fractional integro-differential equation with mixed nonlocal conditions with respect to another function. AIMS Mathematics, 2022, 7(6): 9549-9576. doi: 10.3934/math.2022531 |
[8] | Fan Wan, Xiping Liu, Mei Jia . Ulam-Hyers stability for conformable fractional integro-differential impulsive equations with the antiperiodic boundary conditions. AIMS Mathematics, 2022, 7(4): 6066-6083. doi: 10.3934/math.2022338 |
[9] | Abdulwasea Alkhazzan, Wadhah Al-Sadi, Varaporn Wattanakejorn, Hasib Khan, Thanin Sitthiwirattham, Sina Etemad, Shahram Rezapour . A new study on the existence and stability to a system of coupled higher-order nonlinear BVP of hybrid FDEs under the $ p $-Laplacian operator. AIMS Mathematics, 2022, 7(8): 14187-14207. doi: 10.3934/math.2022782 |
[10] | Hasanen A. Hammad, Hassen Aydi, Manuel De la Sen . The existence and stability results of multi-order boundary value problems involving Riemann-Liouville fractional operators. AIMS Mathematics, 2023, 8(5): 11325-11349. doi: 10.3934/math.2023574 |
In this work, the existence of solutions for nonlinear hybrid fractional integro-differential equations involving generalized proportional fractional (GPF) derivative of Caputo-Liouville-type and multi-term of GPF integrals of Reimann-Liouville type with Dirichlet boundary conditions is investigated. The analysis is accomplished with the aid of the Dhage's fixed point theorem with three operators and the lower regularized incomplete gamma function. Further, the uniqueness of solutions and their Ulam-Hyers-Rassias stability to a special case of the suggested hybrid problem are discussed. For the sake of corroborating the obtained results, an illustrative example is presented.
Fractional calculus has received a lot of attention in the last decades [1]. This type of calculus has become an important area of research due to its intensive development and diverse applications with several complex problems of everyday life have been modeled by differential equations and integro-differential equations of fractional order [2,3,4,5]. Such applications can be found in variety of fields like economics, physics, chemistry, astronomy, thermal acoustic engineering, biology, probability theory, control theory, viscous fluid dynamics, signal processing, electromagnetism, robotics, anomalous diffusion, potential theory and electrical statistics, etc. We refer the reader to the papers [6,7,8] and the references cited therein for more details.
Many authors have taken into consideration the numerical solutions of boundary value problems or initial value problems of ordinary/partial value problems (see [9,10] and the references mentioned). However, studying the qualitative properties of differential equations has still had priority.
In mathematical analysis, boundary conditions are constraints on the solutions of ordinary or partial differential equations in a given domain. There is a large number of possible boundary conditions depending on the formulation of the number of variables involved and thus depending on the nature of the differential equation. Boundary value problems of differential equations of fractional order have been extensively studied. In many works, the authors presented some excellent results about the existence of solutions and utilized the fixed point theorems for this purpose; for example, see [11,12,13,14,15,16]. Concerning the fractional integro-differential equations, please refer to [17,18,19,20,21]. On the other hand, some researchers were interested in studying the stability of solutions to fractional differential equations. One of the most important types of stability that played a noticeable role in the development of fractional differential equations is the so-called Ulam-Hyers stability. Ulam-Hyers stability analysis was recognized as a simple method of investigation of the solutions of a differential equation. We refer the readers to [22,23,24,25,26,27] and the references contained therein.
Some basic theory for the hybrid differential (or hybrid integro-differential) equations was discussed in some articles; for instance, in [28], Dhage and Lakshmikantham discussed the existence and the uniqueness of solutions for the following classical Cauchy problem of hybrid differential equations
$ \begin{equation} \left\{ \begin{array}{l} \frac{d}{dt}\left( \frac{u\left( t\right) }{f(t, u(t))}\right) = g\left( t, u\left( t\right) \right) , \ t\in [0, T], \\ u(0) = u_{0} \end{array} \right. \end{equation} $ | (1.1) |
where $ g\in C([0, T]\times \mathbb{R}, \mathbb{R}) $ and$ \ f\in C([0, T]\times \mathbb{R}, \mathbb{R} \backslash \{0\}). $
Herzallah and Baleanu [29] discussed the existence of solutions and some fundamental differential inequalities for the following Cauchy problems of hybrid differential equations involving Caputo-Liouville derivative.
$ \begin{equation} \left\{ \begin{array}{l} ^{C}D_{0^{+}}^{\beta }\left( \frac{u\left( t\right) }{f(t, u(t))}\right) = g\left( t, u\left( t\right) \right) , \\ u(0) = u_{0} \end{array} \right. \end{equation} $ | (1.2) |
and
$ \begin{equation} \left\{ \begin{array}{l} ^{C}D_{0^{+}}^{\beta }\left( u\left( t\right) -h(t, u(t))\right) = g\left( t, u\left( t\right) \right) , \\ u(0) = u_{0}, \end{array} \right. \end{equation} $ | (1.3) |
where $ t\in [0, T], $ $ 0 < \beta \leq 1, $ $ g\in C([0, T]\times \mathbb{R}, \mathbb{R}) $ and$ \ f, h\in C([0, T]\times \mathbb{R}, \mathbb{R} \backslash \{0\}). $
Bashir et al. [30] investigated the existence of solutions for a nonlocal boundary value problem of hybrid fractional integro-differential inclusions given by
$ \begin{equation} \left\{ \begin{array}{l} ^{C}D_{0^{+}}^{\beta }\left( \frac{u\left( t\right) -\sum\nolimits_{i = 1}^{i = m}I_{0^{+}}^{\alpha _{i}}h_{i}(t, u(t))}{f(t, u(t))}\right) \in g\left( t, u\left( t\right) \right) , \ t\in [0, 1], \\ u(0) = \mu (u), \text{ }u(1) = A, \end{array} \right. \end{equation} $ | (1.4) |
where where $ ^{C}D_{0^{+}}^{\beta } $ denotes the Caputo-Liouville fractional derivative of order $ \beta \in (1, 2] $, $ I_{0^{+}}^{\alpha _{i}} $ denotes the Reimann-Liouville fractional integral of order$ \ \alpha _{i} $ and $ h_{i}\in C([0, T]\times \mathbb{R}, \mathbb{R}), $ $ i = 1, ..., m. $
In [31], Dhage proved the existence and approximations of the solutions for the following initial value problems of nonlinear hybrid fractional integro-differential equations
$ \begin{equation} \left\{ \begin{array}{l} ^{c}D_{0^{+}}^{\beta }\left( \frac{u\left( t\right) -I_{0^{+}}^{\alpha }h(t, u(t))}{f(t, u(t))}\right) = g\left( t, u\left( t\right) , \int_{0}^{t}k(s, u(s))ds\right) , \ t\in [0, T], \\ u(0) = u_{0}, \end{array} \right. \end{equation} $ | (1.5) |
where $ 0 < \beta \leq 1, $ $ \alpha > 0 $ and $ h, k:[0, T]\times \mathbb{R} \longrightarrow \mathbb{R}, $ $ f $ $ :[0, T]\times \mathbb{R} \longrightarrow \mathbb{R} \backslash \{0\}, $ $ g:[0, T]\times \mathbb{R} \times \mathbb{R} \longrightarrow \mathbb{R} $ are given functions.
Motivated by the above mentioned works, in this article, we handle the existence of solutions for the following hybrid proportional fractional integro-differential equations with Dirichlet boundary conditions
$ \begin{equation} \left\{ \begin{array}{l} _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma }\left( \frac{u\left( t\right) -\sum\nolimits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))}{ f(t, u(t))}\right) = g\left( t, u\left( t\right) \right) , \ t\in [0, T], \\ u(0) = u_{0}, u(T) = u_{T}, \end{array} \right. \end{equation} $ | (1.6) |
where $ \sigma \in (0, 1], $ $ 1 < \beta \leq 2, $ $ \alpha _{i} > 0, (i = 1, ..., m), m \in \mathbb{N}^{\ast}, $ $ u_{0}, u_{T} \in \mathbb{R} $ and $ \ _{p}^{\mathcal{ C}}\mathfrak{D}_{0^{+}}^{\beta, \sigma } $ represents the Caputo-Liouville proportional fractional derivative of order $ \beta $ and $ J_{0^{+}}^{\alpha _{i}, \sigma } $ denotes the Reimann-Liouville proportional fractional integral of order $ \alpha _{i} $, and $ h_{i}:[0, T]\times \mathbb{R} \longrightarrow \mathbb{R}, $ $ f $ $ :[0, T]\times \mathbb{R} \longrightarrow \mathbb{R} \backslash \{0\}, g:[0, T]\times \mathbb{R} \longrightarrow \mathbb{ R} $ are given functions. We discuss existence of solutions by applying the Dhage's fixed point and benefiting from the incomplete Gamma function and its properties that were presented by Laadjal et al. [32]. On the top of this, we discuss the uniqueness of these solutions and their Ulam-Hyers-Rassias stability for the case $ f(t, u(t)) = 1 $ for all $ t\in [0, T] $ of the hybrid problem (1.6).
We note that, we deduce the following main cases from problem (1.6).
Case 1. If $ \sigma = 1. $ The problem (1.6) reduces to a hybrid integro-differential equations involving the usual Caputo-Liouville fractional derivative with the usual Reimann-Liouville fractional integral.
Case 2. If $ f\left(t, u\left(t\right) \right) = 1 $ and $ h_{i}\left(t, u\left(t\right) \right) = 0, $ $ (i = 1, ..., m) $ for all $ \left(t, u\left(t\right) \right) \in \lbrack 0, T]\times \mathbb{R}, $ we get the following Caputo-Liouville proportional fractional boundary value problem
$ \begin{equation} \left\{ \begin{array}{l} _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma }u\left( t\right) = g\left( t, u\left( t\right) \right) , t\in \lbrack 0, T], \\ u(0) = u_{0}, u(T) = u_{T}, \end{array} \right. \end{equation} $ | (1.7) |
Case 3. If $ \sigma = 1, f\left(t, u\left(t\right) \right) = 1, $ and$ \ h_{i}\left(t, u\left(t\right) \right) = 0, $ $ (i = 1, ..., m) $ for all $ \left(t, u\left(t\right) \right) \in \lbrack 0, T]\times \mathbb{R}, $ we get the following Caputo-Liouville fractional boundary value problem
$ \begin{equation} \left\{ \begin{array}{l} ^{C}D_{0^{+}}^{\beta }u\left( t\right) = g\left( t, u\left( t\right) \right) , t\in \lbrack 0, T], \\ u(0) = u_{0}, u(T) = u_{T}. \end{array} \right. \end{equation} $ | (1.8) |
The paper is organized as follows: Next section presents some definitions and some properties needed in next sections. Section 3 proves the existence of solutions for the given problem. Section 4 investigates the uniqueness result of solutions for the given problem (with $ f(t, u(t)) = 1 $). Section 5 discusses the Ulam-Hyers-Rassias stability results. Lastly, section 6 provides an example to clarify the results obtained.
In this section, we present some definitions and properties associated with the fractional calculus [33,34] and the incomplete gamma function [32] that are helpful for our discussion.
Definition 4 ([33]). Let $ \beta \geq 0. $ The left fractional integral of Reimann-Liouville type of the function $ \Theta \in L^{1}([0, T], \mathbb{R}) $ is defined by $ I_{a^{+}}^{0}\Theta\left(t\right) = \Theta\left(t\right) $ and
$ \begin{equation} I_{a^{+}}^{\beta }\Theta\left( t\right) = \frac{1}{\Gamma \left( \beta \right) } \int_{a}^{t}\left( t-\rho \right) ^{\beta -1}\Theta\left( \rho \right) d\rho , \ { for } \ \beta > 0, \end{equation} $ | (2.1) |
where $ t\in [a, b] $.
Definition 5 ([33]). Let $ \beta \geq 0. $ The left fractional derivative of Caputo-Liouville type of the function $ \Theta \in C^{n}([0, T], \mathbb{R}) $ is defined by $ ^{C}D_{a^{+}}^{0}\Theta\left(t\right) = \Theta\left(t\right) $ and
$ \begin{eqnarray} ^{C}D_{a^{+}}^{\beta }\Theta\left( t\right) & = &I_{a^{+}}^{n-\beta }\left( D^{n}\Theta\right) (t) \\ & = &\frac{1}{\Gamma \left( n-\beta \right) }\int_{a}^{t}\left( t-\rho \right) ^{n-\beta -1}D^{n}\Theta\left( \rho \right) d\rho \ { for } \ \beta > 0, \end{eqnarray} $ | (2.2) |
where $ n-1 < \beta \leq n $, $ n\in \mathbb{N}. $
Definition 6 ([34]). Let $ \sigma \in (0, 1], $ and $ \beta \geq 0. $ The left generalized proportional fractional integral of Reimann-Liouville type of the function $ \Theta \in L^{1}([0, T], \mathbb{R}) $ is defined by $ J_{a^{+}}^{0, \sigma }\Theta\left(t\right) = \Theta\left(t\right) $ and
$ \begin{equation} J_{a^{+}}^{\beta , \sigma }\Theta\left( t\right) = \frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{a}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{ \sigma -1}{\sigma }(t-\rho )}\Theta\left( \rho \right) d\rho , \ { for } \ \beta > 0, \end{equation} $ | (2.3) |
where $ t\in [a, b] $.
Definition 7 ([34]). Let $ \sigma \in (0, 1], $ and $ \beta \geq 0. $ The left generalized proportional fractional derivative of Caputo-Liouville type of the function $ \Theta \in C^{n}([0, T], \mathbb{R}) $ is defined by $ _{p}^{\mathcal{C}} \mathfrak{D} _{a^{+}}^{0, \sigma }\Theta\left(t\right) = \Theta\left(t\right) $ and
$ \begin{eqnarray} _{p}^{\mathcal{C}}\mathfrak{D}_{a^{+}}^{\beta , \sigma }\Theta\left( t\right) & = &J_{a^{+}}^{n-\beta , \sigma }\left( D^{n, \sigma }\Theta\right) (t) \\ & = &\frac{1}{\sigma ^{\beta }\Gamma \left( n-\beta \right) } \int_{a}^{t}\left( t-\rho \right) ^{n-\beta -1}e^{\frac{\sigma -1}{\sigma } (t-\rho )}\left( D^{n, \sigma }\Theta\right) \left( \rho \right) d\rho \ { for } \ \beta > 0, \end{eqnarray} $ | (2.4) |
where $ n-1 < \beta \leq n $, $ n\in \mathbb{N}. $ and ($ D^{1, \sigma }\Theta)(t) = (D^{\sigma }\Theta)(t) = (1-\sigma)\Theta (t)+\sigma \Theta ^{\prime }(t), $ and
$ \begin{equation} (D^{n, \sigma }\Theta )(t) = \left\{ \begin{array}{l} \Theta (t), \ \ \ { for} \ \ n = 0, \\ \\ \underset{n\ {-times}}{(\underbrace{D^{\sigma }D^{\sigma }\cdots D^{\sigma }}}\Theta )(t), \ \ \ { for}\ \ n\geq 1. \end{array} \right. \end{equation} $ | (2.5) |
Remark 8. Note that, for $ \sigma = 1, $ Definitions 6 and 7 reduce to the usual definitions of Riemann-Liouville fractional integral and Caputo-Liouville fractional derivative, respetively.
Proposition 9 ([34]). Let $ \sigma \in (0, 1], $ $ \alpha > 0 $ and $ \beta > 0 $ with $ n-1 < \beta \leq n $, and $ \Theta \in L^{1}([0, T], \mathbb{R}), $ we have the following properties.
$ \begin{equation} J_{a^{+}}^{\beta , \sigma }\left( J_{a^{+}}^{\alpha , \sigma }\Theta\right) (t) = J_{a^{+}}^{\alpha , \sigma }\left( J_{a^{+}}^{\beta , \sigma }\Theta\right) (t) = J_{a^{+}}^{\beta +\alpha , \sigma }\Theta (t); \end{equation} $ | (2.6) |
$ \begin{equation} _{p}^{\mathcal{C}}\mathfrak{D}_{a^{+}}^{\beta , \sigma }\left( J_{a^{+}}^{\beta , \sigma }\Theta\right) \left( t\right) = \Theta (t); \end{equation} $ | (2.7) |
$ \begin{equation} J_{a^{+}}^{\beta , \sigma }\left( _{p}^{\mathcal{C}}\mathfrak{D} _{a^{+}}^{\beta , \sigma }\Theta\right) \left( t\right) = \Theta\left( t\right) -\sum\limits_{k = 0}^{n-1}c_{k}\left( t-a\right) ^{k}e^{\frac{\sigma -1}{ \sigma } (t-a)}, \mathit{\text{(here}} \ \Theta \in C^{n}([0, T], \mathbb{R})\mathit{\text{), }} \end{equation} $ | (2.8) |
where $ c_{k} = \frac{\left(D^{k, \sigma }\Theta\right) \left(a\right) }{ \sigma ^{k}k!}. $ In particular, when $ 1 < \beta \leq 2 $ and $ a = 0, $ we have
$ \begin{equation} J_{0^{+}}^{\beta , \sigma }\left( _{p}^{\mathcal{C}}\mathfrak{D} _{0^{+}}^{\beta , \sigma }\Theta\right) \left( t\right) = \Theta\left( t\right) -c_{0}e^{ \frac{\sigma -1}{\sigma }t}-c_{1}te^{\frac{\sigma -1}{ \sigma }t}. \end{equation} $ | (2.9) |
Definition 10 ([32,35,36]). Let $ \beta \in $ $ (\Re (\beta) > 0). $
The upper incomplete gamma function is defined by
$ \begin{equation} \Gamma (\beta , t) = \int_{t}^{+\infty }x^{\beta -1}e^{-x}dx, \ \ t\geq 0. \end{equation} $ | (2.10) |
The lower incomplete gamma function is defined by
$ \begin{equation} \gamma (\beta , t) = \int_{0}^{t}x^{\beta -1}e^{-x}dx, \ \ t\geq 0. \end{equation} $ | (2.11) |
The upper regularized incomplete gamma function is defined by
$ \begin{equation} Q(\beta , t) = \frac{\Gamma (\beta , t)}{\Gamma (\beta )}. \end{equation} $ | (2.12) |
The lower regularized incomplete gamma function is defined by
$ \begin{equation} P(\beta , t) = 1-Q(\beta , t) = \frac{\gamma (\beta , t)}{\Gamma (\beta )}. \end{equation} $ | (2.13) |
Remark 11 ([32]). Let $ \varsigma \in \mathbb{R}^{+} $ and $ \beta \in \mathbb{C} $, where $ (\Re (\beta) > 0), $ It is clear that $ P(\beta, \varsigma (t-a)) $ is a non-decreasing function with respect to$ \ t\in [a, b]. $ Moreover
$ \begin{equation} P(\beta , \varsigma (t-a))\in [0, 1] \ { for \ all } \ t\geq a; \end{equation} $ | (2.14) |
$ \begin{equation} \underset{t\in [a, b]}{\max }P(\beta , \varsigma (t-a)) = \left. P(\beta , \varsigma (t-a))\right\vert _{t = b} = P(\beta , \varsigma (b-a)); \end{equation} $ | (2.15) |
$ \begin{equation} \underset{t\in [a, b]}{\min }P(\beta , \varsigma (t-a)) = \left. P(\beta , \varsigma (t-a))\right\vert _{t = a} = 0. \end{equation} $ | (2.16) |
Lemma 12 ([32,37]). Let $ \sigma \in (0, 1], $ and $ \omega > 0. $ Then
$ \begin{equation} \left( J_{a^{+}}^{\omega , \sigma }1\right) \left( t\right) = \left\{ \begin{array}{l} \frac{P(\omega , \frac{1-\sigma }{\sigma }(t-a))}{\left( 1-\sigma \right) ^{\omega }}, \qquad \qquad \mathit{\text{ for }}\sigma \in (0, 1), \\ \left( I_{a^{+}}^{\omega }1\right) \left( t\right) = \frac{(t-a)^{\omega }}{ \Gamma (\omega +1)}, \qquad \mathit{\text{ for }}\sigma = 1, \end{array} \right. \end{equation} $ | (2.17) |
where $ t\in [a, b] $ and the function $ P $ is defined by Eq (2.13). Moreover
$ \begin{equation} \underset{\sigma \rightarrow 1}{\lim }\frac{P(\omega , \frac{1-\sigma }{ \sigma }(t-a))}{\left( 1-\sigma \right) ^{\omega }} = \left( I_{a^{+}}^{\omega }1\right) \left( t\right) = \frac{(t-a)^{\omega }}{\Gamma (\omega +1)} \end{equation} $ | (2.18) |
and
$ \begin{equation} \underset{t\in [a, b]}{\max }\left( \underset{\sigma \rightarrow 1}{ \lim } \frac{P(\omega , \frac{1-\sigma }{\sigma }(t-a))}{\left( 1-\sigma \right) ^{\omega }}\right) = \frac{(b-a)^{\omega }}{\Gamma (\omega +1)}. \end{equation} $ | (2.19) |
Lemma 13 ([32,37]). Let $ \sigma \in (0, 1), $ $ t_{1}, t_{2}\in [a, b] $ $ (t_{1}\leq t_{2}) $ and $ \omega > 0. $ Then
$ \begin{equation} \int_{t_{1}}^{t_{2}}\left( b-\rho \right) ^{\omega -1}e^{\frac{\sigma -1}{ \sigma }(b-\rho )}d\rho = \frac{\sigma ^{\omega }\Gamma \left( \omega \right) }{\left( 1-\sigma \right) ^{\omega }}\left[P(\omega , \frac{1-\sigma }{ \sigma }(b-t_{1}))-P(\omega , \frac{1-\sigma }{\sigma }(b-t_{2}))\right] , \end{equation} $ | (2.20) |
where the function $ P $ is defined by Eq (2.13).
Lemma 14 ([32,37]). Let $ \sigma \in (0, 1], $ $ \beta > 0 $ and$ \ a\leq \rho \leq t_{1} < t_{2}\leq b. $ Then
$ \begin{equation} \underset{t_{2}\rightarrow t_{1}}{\lim }\int_{a}^{t_{1}}\left\vert \left( t_{2}-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t_{2}-\rho )}-\left( t_{1}-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma } (t_{1}-\rho )}\right\vert d\rho = 0. \end{equation} $ | (2.21) |
Theorem 15 (Dhage [38]). Let $ \mho $ be a non-empty, closed convex and bounded subset of a Banach algebra $ \Sigma $ and let $ A, C:\Sigma \rightarrow \Sigma $ and $ B:\mho\rightarrow \Sigma $ be three operators satisfying the following conditions.
(1) $ A $ and $ C $ are Lipschitz with Lipschitz constants $ \widetilde{A} $ and $ \widetilde{C} $, respectively
(2) $ B $ is completely continuous
(3) $ \ \widetilde{A}M_{B}+\widetilde{C} < 1 $, where $ M_{B} = \left\Vert B(\mho)\right\Vert = \sup \left\{ \parallel Bx\parallel :x\in \mho\right\} $
(4) $ AxBy+Cx = x\Rightarrow x\in \mho $ for all $ y\in \mho $.
Then, the operator equation
$ \begin{equation} AxBx+Cx = x \end{equation} $ | (2.22) |
has a solution in $ \mho. $
In this section, we provide our essential findings concerning the problem defined by (1.6) and then derive the existence results for the special cases 1, 2 and 3.
Lemma 16. Let $ \sigma \in (0, 1], $ $ 1 < \beta \leq 2, $ $ \alpha _{i} > 0, (i = 1, ..., m). $ For $ \Theta \in C([0, T], \mathbb{R}), $ the solution of
$ \begin{equation} \left\{ \begin{array}{l} _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma }\left( \frac{ u\left( t\right) -\sum\nolimits_{i = 1}^{i = m}J^{\alpha _{i}, \sigma }h_{i}(t, u(t))}{ f(t, u(t))}\right) = \Theta \left( t\right) \\ u(0) = u_{0}, u(T) = u_{T}, \end{array} \right. \end{equation} $ | (3.1) |
is equivalent to the integral equation
$ \begin{eqnarray} u(t) & = &f(t, u(t))\Big[\frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T}\right) e^{ \frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &-&\frac{te^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})Te^{\frac{\sigma -1}{ \sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{ \frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho \\ &-&\frac{te^{\frac{\sigma -1}{\sigma }t}}{Te^{\frac{\sigma -1}{\sigma }T}} \frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\Theta \left( \rho \right) d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\Theta \left( \rho \right) d\rho \Big] \\ &+&\underset{i = 1}{\overset{m}{\sum }}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho . \end{eqnarray} $ | (3.2) |
Proof. Applying the operator $ J_{0^{+}}^{\beta, \sigma } $ on both sides of the equation in Eq (3.1), we get
$ \begin{equation} \frac{u\left( t\right) -\sum\nolimits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))}{f(t, u(t))}-c_{0}e^{\frac{\sigma -1}{\sigma }t}-c_{1}te^{ \frac{\sigma -1}{\sigma }t} = J_{0^{+}}^{\beta , \sigma }\Theta (t). \end{equation} $ | (3.3) |
From the boundary conditions $ u(0) = u_{0} $ and $ u(T) = u_{T} $ we get
$ \begin{equation} c_{0} = \frac{u_{0}}{f(0, u_{0})}, \end{equation} $ | (3.4) |
and
$ \begin{equation} c_{1} = \frac{u_{T}-\sum\nolimits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(., u(.))(T)}{f(T, u_{T})Te^{\frac{\sigma -1}{\sigma }T}}-\frac{u_{0}}{ f(0, u_{0})T}-\frac{1}{Te^{\frac{\sigma -1}{\sigma }T}}J_{0^{+}}^{\beta , \sigma }\Theta (T). \end{equation} $ | (3.5) |
Substituting the values of $ c_{0}\ $and $ c_{1} $ in Eq (3.3), we obtain
$ \begin{eqnarray} u\left( t\right) & = &f(t, u(t))\Big[\frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T }\right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{ \sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &&-\frac{te^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})Te^{\frac{\sigma -1}{ \sigma }T}}\sum\limits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(\cdot , u(\cdot ))(T) \\ &&-\frac{te^{\frac{\sigma -1}{\sigma }t}}{Te^{\frac{\sigma -1}{\sigma }T}} J^{\beta , \sigma }\Theta (T)+J_{0^{+}}^{\beta , \sigma }\Theta (t)\Big] +\sum\limits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t)). \end{eqnarray} $ |
Inversely, it is obvious that if $ u(t) $ satisfies Eq (3.2), then Eq (3.1) holds. The proof is complete.
Now, consider the Banach algebra space defined by $ \Sigma = C([0, T], \mathbb{R}), $ with the norm $ \left\Vert u\right\Vert = \underset{0\leq t\leq T}{\sup } \left\vert u(t)\right\vert. $ Then, we define the multiplication in $ \Sigma $ by $ (uv)(t) = u(t)v(t) $ for all $ u, v\in \Sigma $ and the operator $ \Phi :\Sigma \rightarrow \Sigma $ $ \ $by
$ \begin{eqnarray} \Phi u(t) & = &f(t, u(t))\Big[\frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T} \right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \Big] \\ &+&\underset{i = 1}{\overset{i = m}{\sum }}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho . \end{eqnarray} $ | (3.6) |
Remark 17. The problem described by Eq (1.6) has a solution $ u\in \Sigma $ if only if$ \ u $ is fixed point of the operator $ \Phi $.
To prove the existence, we need the following assumption:
$ (H_{1}) $ Assume that $ g:[0, T]\times \mathbb{R} \rightarrow \mathbb{R} $ is continuous function, there exists a continuous function $ \widetilde{g} $ from $ [0, T] $ to $ \mathbb{R}_{+} $ and a continuous nondecreasing function $ \psi $ from $ \mathbb{R}_{+} $ to $ \mathbb{R}_{+}\backslash \{0\} $ such that $ \left\vert g(t, u(t))\right\vert \leq \widetilde{g}(t)\psi (\left\Vert u\right\Vert), $ for all $ t\in [0, T] $ with $ \underset{0\leq t\leq T}{\sup } \widetilde{g}(t) = \left\Vert \widetilde{g} \right\Vert. $
$ (H_{2}) $ Assume that $ h_i:[0, T]\times \mathbb{R}\rightarrow \mathbb{R} $, $ i = 1, ..., 4, $ are continuous functions, there exist $ L_{h_{i}} > 0, $ $ i = 1, ..., m $ such that $ \left\vert h_{i}(t, \lambda_1)-h_{i}(t, \lambda_2)\right\vert \leq L_{h_{i}}\left\vert \lambda_1-\lambda_2\right\vert, $ for all $ (t, \lambda_1, \lambda_2)\in [0, T]\times \mathbb{R}\times \mathbb{R} $ and $ \left\vert h_{i}(t, 0)\right\vert \leq \theta _{i}(t), $ where $ \theta _{i}(t) $ are positive and continuous functions on $ [0, T] $ for $ i = 1, ..., m $ with $ \underset{ 0\leq t\leq T}{\sup }\theta _{i}(t) = \left\Vert \theta _{i} \right\Vert. $
$ (H_{3}) $ Assume that there exists $ L_{f} > 0 $ such that $ \left\vert f(t, \lambda _{1})-f(t, \lambda _{2})\right\vert \leq L_{f}\left\vert \lambda _{1}-\lambda _{2}\right\vert, $ for all $ (t, \lambda _{1}, \lambda _{2})\in [0, T]\times \mathbb{R} \times \mathbb{R}, $ and $ \left\vert f(t, 0)\right\vert \leq \zeta (t), $ where $ \zeta (t) $ is positive and continuous function on $ [0, T] $ with $ \underset{0\leq t\leq T}{\sup }\zeta (t) = \left\Vert \zeta \right\Vert. $
Now, we present the following existence theorem for the problem described in Eq (1.6) using the Dhage's fixed point theorem.
Theorem 18. Let $ \sigma \in (0, 1) $ and the hypotheses $ (H_{1}), (H_{2}) $ and $ (H_{3}) $ hold. If there exists a real number $ R > 0 $ such that
$ \begin{equation} R\geq \frac{\varrho \left\Vert \zeta \right\Vert +\sum\nolimits_{i = 1}^{i = m}\frac{ \left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{ \left( 1-\sigma \right) ^{\alpha _{i}}}}{1-\varrho L_{f}-\sum\nolimits_{i = 1}^{i = m} \frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}}, \end{equation} $ | (3.7) |
where $ P $ denotes the lower regularized incomplete gamma function with
$ \begin{equation} \varrho L_{f}+\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{ \sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} < 1, \end{equation} $ | (3.8) |
and
$ \begin{eqnarray} \varrho & = &\frac{\left\vert u_{0}\right\vert }{\left\vert f(0, u_{0})\right\vert }+\frac{1}{\left\vert f(T, u_{T})\right\vert e^{\frac{ \sigma -1}{\sigma }T}}\Big[\left\vert u_{T}\right\vert +\sum\limits_{i = 1}^{i = m} \frac{\left( L_{h_{i}}R+\left\Vert \theta _{i}\right\Vert \right) P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} \Big] \\ &&+\left( \frac{1}{e^{\frac{\sigma -1}{\sigma }T}}+1\right) \frac{\left\Vert \widetilde{g}\right\Vert \psi (R)P(\beta , \frac{1-\sigma }{\sigma }T)}{ \left( 1-\sigma \right) ^{\beta }}. \end{eqnarray} $ | (3.9) |
Then, the problem described in Eq (1.6) has at least one solution on $ [0, T]. $
Proof. Consider $ \mho = \{u\in \Sigma, $ s.t. $ \left\Vert u\right\Vert \leq R\} $ and let
$ \begin{equation} \Phi u = AuBu+Cu, \end{equation} $ | (3.10) |
where $ A, C:\Sigma \rightarrow \Sigma $ and $ B:\mho \rightarrow \Sigma $ are three operators defined by
$ \begin{equation} Au\left( t\right) = f(t, u(t)), \end{equation} $ | (3.11) |
$ \begin{eqnarray} Bu(t) & = &\frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T}\right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{ f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \end{eqnarray} $ | (3.12) |
and
$ \begin{equation} Cu(t) = \sum\limits_{i = 1}^{i = m}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{ \sigma -1}{\sigma }(t-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho . \end{equation} $ | (3.13) |
Claim (1). $ A $ and $ C $ are Lipschitz with Lipschitz constants $ \widetilde{A} $ and $ \widetilde{C} $, respectively.
Let $ u, v\in \Sigma $, for all $ t\in [0, T] $ we have
$ \begin{equation*} \left\vert Au(t)-Av(t)\right\vert = \left\vert f(t, u(t))-f(t, v(t))\right\vert \leq L_{f}\left\Vert u-v\right\Vert , \end{equation*} $ |
which yields
$ \begin{equation*} \left\Vert Au-Av\right\Vert \leq \widetilde{A}\left\Vert u-v\right\Vert , \text{ where }\widetilde{A} = L_{f}. \end{equation*} $ |
Next, we have
$ \begin{eqnarray} \left\vert Cu(t)-Cv(t)\right\vert &\leq &\sum\limits_{i = 1}^{i = m}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\left\vert h_{i}(\rho , u\left( \rho \right) )-h_{i}(\rho , v\left( \rho \right) )\right\vert d\rho \\ &\leq &\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{ \frac{\sigma -1}{\sigma }(t-\rho )}d\rho \left\Vert u-v\right\Vert \\ & = &\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }t)}{ \left( 1-\sigma \right) ^{\alpha _{i}}}\left\Vert u-v\right\Vert \\ &\leq &\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma } T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}\left\Vert u-v\right\Vert , \end{eqnarray} $ |
which yields
$ \begin{equation} \left\Vert Cu-Cv\right\Vert \leq \widetilde{C}\left\Vert u-v\right\Vert , \text{ where }\widetilde{C} = \sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}. \end{equation} $ | (3.14) |
Hence $ A $ and $ C $ are Lipschitz with Lipschitz constants $ \widetilde{A} $ and $ \widetilde{C} $, respectively.
Claim (2). $ B $ is completely continuous.
Step 1. We will show that the operator $ B:\mho\rightarrow \Sigma $ is uniformly bounded.
For any $ u\in \mho $, we have
$ \begin{eqnarray} \left\vert Bu(t)\right\vert & = &\frac{\left\vert u_{0}\right\vert }{ \left\vert f(0, u_{0})\right\vert }\left( 1-\frac{t}{T}\right) e^{\frac{ \sigma -1}{\sigma }t}+\frac{\left\vert u_{T}\right\vert \frac{t}{T}e^{\frac{ \sigma -1}{\sigma }t}}{\left\vert f(T, u_{T})\right\vert e^{\frac{\sigma -1}{ \sigma }T}} \\ &+&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\left\vert f(T, u_{T})\right\vert e^{\frac{\sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{ \int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\left\vert h_{i}(\rho , u\left( \rho \right) )\right\vert d\rho }{ \sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &+&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\left\vert g\left( \rho , u\left( \rho \right) \right) \right\vert d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\left\vert g\left( \rho , u\left( \rho \right) \right) \right\vert d\rho . \end{eqnarray} $ |
For $ u\in \mho $ and $ t\in [0, T] $, using $ (H_{3}) $ we get
$ \begin{eqnarray} \left\vert h_{i}(t, u\left( t\right) )\right\vert & = &\left\vert h_{i}(t, u\left( t\right) )-h_{i}(t, 0)+h_{i}(t, 0)\right\vert \\ &\leq &\left\vert h_{i}(t, u\left( t\right) )-h_{i}(t, 0)\right\vert +\left\vert h_{i}(t, 0)\right\vert \\ &\leq &L_{h_{i}}R+\left\Vert \theta _{i}\right\Vert . \end{eqnarray} $ |
Then, using Lemma 13, we obtain
$ \begin{eqnarray} \left\vert Bu(t)\right\vert &\leq &\frac{\left\vert u_{0}\right\vert }{ \left\vert f(0, u_{0})\right\vert }+\frac{\left\vert u_{T}\right\vert }{ \left\vert f(T, u_{T})\right\vert e^{\frac{\sigma -1}{\sigma }T}}+\frac{1}{ \left\vert f(T, u_{T})\right\vert e^{\frac{\sigma -1}{\sigma }T}} \sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}R+\left\Vert \theta _{i}\right\Vert \right) P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} \\ &+&\frac{\left\Vert \widetilde{g}\right\Vert \psi (R)P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }e^{\frac{\sigma -1}{\sigma }T} }+\frac{\left\Vert \widetilde{g}\right\Vert \psi (R)P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }} \\ & = &\varrho < +\infty . \end{eqnarray} $ |
Therefore, $ \Vert Bu\Vert < +\infty $, and consequently $ B(\mho) $ is bounded. Hence, $ B $ is uniformly bounded.
Step 2.We show that the continuity of $ B $ on $ \mho. $ Let $ \left(u_{n}\right) _{n\in \mathbb{N}} $ be a sequence where $ \underset{n\rightarrow +\infty }{\lim }u_{n} = u $ with $ u\in \mho $.
For all $ t\in [0, T] $, we have
$ \begin{eqnarray} \lim\limits_{n\rightarrow +\infty }Bu_{n}(t) & = &\frac{u_{0}}{f(0, u_{0})}\left( 1- \frac{t}{T}\right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{ \frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\lim\limits_{n\rightarrow +\infty }\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{ \sigma }(T-\rho )}h_{i}(\rho , u_{n}\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\lim\limits_{n\rightarrow +\infty }\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{ \sigma }(T-\rho )}g\left( \rho , u_{n}\left( \rho \right) \right) d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\lim\limits_{n\rightarrow +\infty }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{ \sigma }(t-\rho )}g\left( \rho , u_{n}\left( \rho \right) \right) d\rho . \end{eqnarray} $ |
Using the Lebesgue dominated convergence theorem, we get
$ \begin{eqnarray} \lim\limits_{n\rightarrow +\infty }Bu_{n}(t) & = &\frac{u_{0}}{f(0, u_{0})}\left( 1- \frac{t}{T}\right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{ \frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\lim\limits_{n\rightarrow +\infty }h_{i}(\rho , u_{n}\left( \rho \right) )d\rho }{ \sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &-&\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\lim\limits_{n\rightarrow +\infty }g\left( \rho , u_{n}\left( \rho \right) \right) d\rho \\ &+&\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\lim\limits_{n\rightarrow +\infty }g\left( \rho , u_{n}\left( \rho \right) \right) d\rho \\ & = &Bu(t), \end{eqnarray} $ |
which yields $ \lim_{n\rightarrow +\infty }\left\Vert Bu_{n}-Bu\right\Vert = 0. $ Therefore, $ B $ is continuous on $ \mho $.
Step 3.We prove that $ B(\mho) $ is equicontinuous.
Let $ t_{1}, t_{2}\in I = [0, T] $ with $ t_{1} > t_{2}. $ Then, for any $ u\in \mho $, we have
$ \begin{eqnarray} \left\vert Bu(t_{2})-Bu(t_{1})\right\vert & = &\Big|\frac{u_{0}}{f(0, u_{0})}\left\{ \left( 1-\frac{t_{2}}{T}\right) e^{ \frac{\sigma -1}{\sigma }t_{2}}-\left( 1-\frac{t_{1}}{T}\right) e^{\frac{ \sigma -1}{\sigma }t_{1}}\right\} +\frac{u_{T}\left\{ \frac{t_{2}}{T}e^{ \frac{\sigma -1}{\sigma }t_{2}}-\frac{t_{1}}{T}e^{\frac{\sigma -1}{\sigma } t_{1}}\right\} }{f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &&-\frac{\left\{ \frac{t_{2}}{T}e^{\frac{\sigma -1}{\sigma }t_{2}}-\frac{ t_{1}}{T}e^{\frac{\sigma -1}{\sigma }t_{1}}\right\} }{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &&-\frac{\left\{ \frac{t_{2}}{T}e^{\frac{\sigma -1}{\sigma }t_{2}}-\frac{ t_{1}}{T}e^{\frac{\sigma -1}{\sigma }t_{1}}\right\} }{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \\ &&+\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t_{2}} \left( t_{2}-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t_{2}-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \\ &&-\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t_{1}} \left( t_{1}-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t_{1}-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \Big| \\ &\leq &\frac{\left\vert u_{0}\right\vert }{\left\vert f(0, u_{0})\right\vert } \left\vert \left( 1-\frac{t_{2}}{T}\right) e^{\frac{\sigma -1}{\sigma } t_{2}}-\left( 1-\frac{t_{1}}{T}\right) e^{\frac{\sigma -1}{\sigma } t_{1}}\right\vert +\frac{\left\vert u_{T}\right\vert \left\vert \frac{t_{2}}{ T}e^{\frac{\sigma -1}{\sigma }t_{2}}-\frac{t_{1}}{T}e^{\frac{\sigma -1}{ \sigma }t_{1}}\right\vert }{\left\vert f(T, u_{T})\right\vert e^{\frac{\sigma -1}{\sigma }T}} \\ &&+\frac{\left\vert \frac{t_{2}}{T}e^{\frac{\sigma -1}{\sigma }t_{2}}-\frac{ t_{1}}{T}e^{\frac{\sigma -1}{\sigma }t_{1}}\right\vert }{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}R+\left\Vert \theta _{i} \right\Vert\right) \int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{ \sigma -1}{\sigma }(T-\rho )}d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &&+\frac{\left\vert \frac{t_{2}}{T}e^{\frac{\sigma -1}{\sigma }t_{2}}-\frac{ t_{1}}{T}e^{\frac{\sigma -1}{\sigma }t_{1}}\right\vert \left\Vert \widetilde{ g} \right\Vert \psi (R)}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{ \frac{\sigma -1}{\sigma }T}} \int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{ \frac{\sigma -1}{\sigma } (T-\rho )}d\rho \\ &&+\frac{\left\Vert \widetilde{g} \right\Vert \psi (R)}{\sigma ^{\beta }\Gamma \left( \beta \right) } \int_{0}^{t_{1}}\left\vert \left( t_{2}-\rho \right) ^{\beta -1}e^{\frac{ \sigma -1}{\sigma }(t_{2}-\rho )}-\left( t_{1}-\rho \right) ^{\beta -1}e^{ \frac{\sigma -1}{\sigma }(t_{1}-\rho )}\right\vert d\rho \\ &&+\frac{\left\Vert \widetilde{g} \right\Vert \psi (R)}{\sigma ^{\beta }\Gamma \left( \beta \right) } \int_{t_{1}}^{t_{2}}\left( t_{2}-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{ \sigma }(t_{2}-\rho )}d\rho . \end{eqnarray} $ |
Using Lemmas 16 and 14, we obtain
$ \begin{equation*} \left\vert Bu(t_{2})-Bu(t_{1})\right\vert \rightarrow 0\text{ as } t_{1}\rightarrow t_{2}. \end{equation*} $ |
Therefore, $ Bu(t) $ is equicontinuous on $ [0, T] $.
Making use of the Arzela-Ascoli theorem, we have $ B(\mho) $ is relatively compact. Thus $ B $ is a compact operator and as a consequence $ B $ is completely continuous.
Claim (3). $ \widetilde{A}M_{B}+\widetilde{C} < 1 $, where $ M_{B} = \sup \left\{ \parallel Bu\parallel :u\in \mho\right\}. $
We have $ M_{B} = \left\Vert B(\mho)\right\Vert = \sup \left\{ \parallel Bu\parallel :u\in \mho\right\} \leq \varrho $ where $ \varrho $ is given by(3.9). So,
$ \begin{equation*} \widetilde{A}M_{B}+\widetilde{C}\leq \varrho L_{f}+\sum\limits_{i = 1}^{i = m}\frac{ L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} < 1. \end{equation*} $ |
Claim (4).$ AuB\overline{u}+Cu = u\Rightarrow u\in \mho $ for all $ \overline{u}\in \mho. $
Let $ \overline{u}\in \mho $ we have
$ \begin{eqnarray} \left\Vert u\right\Vert &\leq &\left\Vert Au\right\Vert \cdot \left\Vert B \overline{u}\right\Vert +\left\Vert Cu\right\Vert \\ & = &\underset{t\in [0, T]}{\sup }\Big[\left\vert f(t, u(t))\right\vert \Big| \frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T}\right) e^{\frac{\sigma -1}{ \sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{ f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &&-\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &&-\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \\ &&+\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}g\left( \rho , u\left( \rho \right) \right) d\rho \Big| \\ &&+\Big|\sum\limits_{i = 1}^{i = m}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{ \sigma -1}{\sigma }(t-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho \Big| \Big] \\ &\leq &\underset{t\in [0, T]}{\sup }\Big[\left( L_{f}\left\Vert u\right\Vert +\left\Vert \zeta \right\Vert \right) \varrho +\sum\limits_{i = 1}^{i = m}\left( L_{h_{i}}\left\Vert u\right\Vert +\left\Vert \theta _{i}\right\Vert \right) \frac{P(\alpha _{i}, \frac{1-\sigma }{\sigma }t)}{\left( 1-\sigma \right) ^{\alpha _{i}}}\Big] \\ & = &\left( L_{f}\left\Vert u\right\Vert +\left\Vert \zeta \right\Vert \right) \varrho +\underset{i = 1}{\overset{m}{\sum }}\left( L_{h_{i}}\left\Vert u\right\Vert +\left\Vert \theta _{i}\right\Vert \right) \frac{P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} \\ & = &\left\Vert u\right\Vert \left( \varrho L_{f}+\underset{i = 1}{\overset{m}{ \sum }}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}\right) +\varrho \left\Vert \zeta \right\Vert +\underset{i = 1}{\overset{m}{\sum }}\frac{\left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}, \end{eqnarray} $ |
which yields
$ \begin{equation*} \left\Vert u\right\Vert \leq \frac{\varrho \left\Vert \zeta \right\Vert +\sum\nolimits_{i = 1}^{i = m}\frac{\left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{ 1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}}{1-\varrho L_{f}-\sum\nolimits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma } T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}}\leq R. \end{equation*} $ |
Therefore, $ u\in \mho $.
Thus, all the conditions of Dhage fixed point theorem are satisfied; hence, the operator $ \Phi $ has a fixed point in $ \mho $. As a result, the proportional fractional boundary value problem declaired in Eq (1.6) has at least one solution on $ [0, T]. $ This completes the proof.
In the following, we present the existence theorem for the given problems in the special cases 1, 2 and 3.
Remark 19. From Lemma 12, in case $ \sigma = 1, $ we can replace the formula $ \frac{P(\omega, \frac{1-\sigma }{\sigma }T)}{\left(1-\sigma \right) ^{\omega }} $ by the formula $ \frac{T^{\omega }}{\Gamma (\omega +1)} $ ($ \omega \in \left\{ \beta, \alpha _{1}, ..., \alpha _{m}\right\} $). Then, by using Theorem 18 we can conclude the existence results of the given problem with usual Caputo fractional derivative.
Corollary 20. Let $ \sigma = 1 $ (i.e., $ _{p}^{\mathcal{C}}\mathfrak{D} _{0^{+}}^{\beta, \sigma } = $ $ ^{C}D_{0^{+}}^{\beta } $) and assume that the hypotheses $ (H_{1}), (H_{2}) $ and $ (H_{3}) $ hold. If there exists a real number $ R > 0 $ such that
$ \begin{equation} R\geq \frac{\widetilde{\varrho }\left\Vert \zeta \right\Vert +\sum\nolimits_{i = 1}^{i = m}\frac{\left\Vert \theta _{i}\right\Vert T^{\alpha _{i}}}{ \Gamma (\alpha _{i}+1)}}{1-\widetilde{\varrho }L_{f}-\sum\nolimits_{i = 1}^{i = m}\frac{ L_{h_{i}}T^{\alpha _{i}}}{\Gamma (\alpha _{i}+1)}}, \end{equation} $ | (3.15) |
where
$ \begin{equation} \widetilde{\varrho }L_{f}+\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}T^{\alpha _{i}}}{ \Gamma (\alpha _{i}+1)} < 1 \end{equation} $ | (3.16) |
and
$ \begin{equation} \widetilde{\varrho } = \frac{\left\vert u_{0}\right\vert }{\left\vert f(0, u_{0})\right\vert }+\frac{1}{\left\vert f(T, u_{T})\right\vert }\Big[ \left\vert u_{T}\right\vert +\sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}R+\left\Vert \theta _{i}\right\Vert \right) T^{\alpha _{i}}}{\Gamma (\alpha _{i}+1)}\Big]+\frac{2\left\Vert \widetilde{g}\right\Vert \psi (R)T^{\beta }}{\Gamma (\beta +1)}. \end{equation} $ | (3.17) |
Then, the identified problem in Eq (1.6) has at least one solution on $ [0, T]. $
Corollary 21. Let $ \sigma \in (0, 1) $ and assume that the hypothesis $ (H_{1}) $ holds. If there exists a real number $ R > 0 $ such that
$ \begin{equation} R\geq \frac{\varrho _{0}}{1-\varrho _{0}}, \end{equation} $ | (3.18) |
where
$ \begin{equation} \varrho _{0} = \left\vert u_{0}\right\vert +\frac{\left\vert u_{T}\right\vert }{e^{\frac{\sigma -1}{\sigma }T}}+\left( \frac{1}{e^{\frac{\sigma -1}{\sigma }T}}+1\right) \frac{\left\Vert \widetilde{g}\right\Vert \psi (R)P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }} < 1. \end{equation} $ | (3.19) |
Then, the problem in (1.7) has at least one solution on $ [0, T]. $
Corollary 22. Let $ \sigma = 1, $ and assume that the hypothesis $ (H_{1}) $ holds. If there exists a real number $ R > 0 $ such that
$ \begin{equation} R\geq \frac{\varrho _{1}}{1-\varrho _{1}}, \end{equation} $ | (3.20) |
where
$ \begin{equation} \varrho _{1} = \left\vert u_{0}\right\vert +\left\vert u_{T}\right\vert +\frac{ 2\left\Vert \widetilde{g}\right\Vert \psi (R)T^{\beta }}{\Gamma (\beta +1)} < 1. \end{equation} $ | (3.21) |
Then, the problem (1.8) has at least one solution on $ [0, T]. $
In this section, we discuss the existence and uniqueness of solution for the following problem.
$ \begin{equation} \left\{ \begin{array}{l} _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma }\left( u\left( t\right) -\sum\nolimits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))\right) = g\left( t, u\left( t\right) \right) , \ t\in [0, T], \\ u(0) = u_{0}, u(T) = u_{T}, \end{array} \right. \end{equation} $ | (4.1) |
Note that we can write the equivalent equation for problem described by Eq (4.1) as follows.
$ \begin{equation} u(t) = B_{1}u(t)+Cu(t): = \Phi _{1}u(t), \end{equation} $ | (4.2) |
where $ B_{1} = B $ (with $ f(0, u_{0}) = 1 = f(T, u_{T}) $) and $ B $ and $ C $ are given by Eq (3.12) and Eq (3.13), respectively.
The following assumption is essential.
$ (H_{4}) $ Assume $ g\in C([0, T]^{2}\times \mathbb{R}, \mathbb{R}) $ and there exists $ L_{g} > 0 $ such that $ \left\vert g(t, \lambda _{1})-g(t, \lambda _{2})\right\vert \leq L_{g}\left\vert \lambda _{1}-\lambda _{2}\right\vert, $ for all $ (t, \lambda _{1}, \lambda _{2})\in [0, T]\times \mathbb{R}\times \mathbb{R} $ and $ \left\vert g(t, 0)\right\vert \leq \Upsilon (t), $ where $ \Upsilon (t) $ is positive and continuous function on $ [0, T] $, with $ \underset {0\leq t\leq T}{\sup }\Upsilon (t) = \left\Vert \Upsilon \right\Vert. $
Theorem 23. Let $ \sigma \in (0, 1) $ and assume that $ (H_{2}) $ and $ (H_{4}) $ are satisfied. Then, the problem described by Eq (4.1) has a unique solution on $ [0, T] $ if
$ \begin{equation} \Delta < 1, \end{equation} $ | (4.3) |
where
$ \begin{equation} \Delta = \left( \frac{L_{g}P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }}+\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}\right) \left( e^{\frac{1-\sigma }{\sigma }T}+1\right) \end{equation} $ | (4.4) |
with $ P $ defined as in Eq (2.13).
Proof. Let us set $ \overline{\mho } = \{u\in \Sigma, $ s.t. $ \left\Vert u\right\Vert \leq r\}, $ where $ r > 0 $ satisfying:
$ \begin{equation*} r\geq \frac{\left\vert u_{0}\right\vert +\left\vert u_{T}\right\vert e^{ \frac{1-\sigma }{\sigma }T}+\overline{\Delta }}{1-\Delta }, \end{equation*} $ |
where $ \Delta $ is given by Eq (4.4) and
$ \begin{equation} \overline{\Delta } = \left( e^{\frac{1-\sigma }{\sigma }T}+1\right) \left( \frac{\left\Vert \Upsilon \right\Vert P(\beta , \frac{1-\sigma }{\sigma }T)}{ \left( 1-\sigma \right) ^{\beta }}+\sum\limits_{i = 1}^{i = m}\frac{\left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}\right) . \end{equation} $ | (4.5) |
We will show that $ \Phi _{1}\overline{\mho }\subset \overline{\mho }. $
For $ u\in \overline{\mho } $ and $ t\in [0, T], $ we have
$ \begin{eqnarray} \left\vert B_{1}u(t)\right\vert &\leq &\left\vert u_{0}\right\vert +\frac{ \left\vert u_{T}\right\vert }{e^{\frac{\sigma -1}{\sigma }T}}+\frac{1}{e^{ \frac{\sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}r+\left\Vert \theta _{i}\right\Vert \right) P(\alpha _{i}, \frac{ 1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} \\ &&+\frac{\left( L_{g}r+\left\Vert \Upsilon \right\Vert \right) P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }e^{\frac{ \sigma -1}{\sigma }T}}+\frac{\left( L_{g}r+\left\Vert \Upsilon \right\Vert \right) P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }}. \end{eqnarray} $ |
On the other hand, we have
$ \begin{equation} \left\Vert Cu\right\Vert \leq \sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}r+\left\Vert \theta _{i}\right\Vert \right) P(\alpha _{i}, \frac{ 1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}. \notag \end{equation} $ |
We procure that
$ \begin{eqnarray} \left\Vert \Phi _{1}u\right\Vert &\leq &\left\Vert B_{1}u\right\Vert +\left\Vert Cu\right\Vert \\ &\leq &\left\vert u_{0}\right\vert +\frac{\left\vert u_{T}\right\vert }{e^{ \frac{\sigma -1}{\sigma }T}}+\left( e^{\frac{1-\sigma }{\sigma }T}+1\right) \sum\limits_{i = 1}^{i = m}\frac{\left( L_{h_{i}}r+\left\Vert \theta _{i}\right\Vert \right) P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} \\ &&+\left( e^{\frac{1-\sigma }{\sigma }T}+1\right) \frac{\left( L_{g}r+\left\Vert \Upsilon \right\Vert \right) P(\beta , \frac{1-\sigma }{ \sigma }T)}{\left( 1-\sigma \right) ^{\beta }} \\ & = &\left\vert u_{0}\right\vert +\left\vert u_{T}\right\vert e^{\frac{ 1-\sigma }{\sigma }T}+\overline{\Delta }+\Delta r \\ &\leq &r, \end{eqnarray} $ |
which implies that $ \Phi _{1}\overline{\mho }\subset \overline{\mho }. $
Next, we show that the operator $ \Phi _{1} $ is a contraction mapping.
For $ u, v\in \Sigma $ and for all $ t\in [0, T], $ we have
$ \begin{eqnarray} \left\vert B_{1}u(t)-B_{1}v(t)\right\vert &\leq &\frac{\frac{t}{T}e^{\frac{ \sigma -1}{\sigma }t}}{e^{\frac{\sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{ \int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\left\vert h_{i}(\rho , u\left( \rho \right) )-h_{i}(\rho , v\left( \rho \right) )\right\vert d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &&+\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\left\vert g\left( \rho , u\left( \rho \right) \right) -g(\rho , v\left( \rho \right) )\right\vert d\rho \\ &&+\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\left\vert g\left( \rho , u\left( \rho \right) \right) -g(\rho , u\left( \rho \right) )\right\vert d\rho \\ &\leq &\frac{1}{e^{\frac{\sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{ L_{h_{i}}\left\Vert u-v\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T) }{\left( 1-\sigma \right) ^{\alpha _{i}}} +\frac{L_{g}\left\Vert u-v\right\Vert P(\beta , \frac{1-\sigma }{\sigma }T) }{\left( 1-\sigma \right) ^{\beta }e^{\frac{\sigma -1}{\sigma }T}} \\ &&+\frac{L_{g}\left\Vert u-v\right\Vert P(\beta , \frac{1-\sigma }{\sigma }t)}{\left( 1-\sigma \right) ^{\beta }} \\ &\leq &\bigg\{\frac{1}{e^{\frac{\sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{ L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}+\frac{L_{g}P(\beta , \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\beta }}\left( e^{\frac{1-\sigma }{\sigma }T}+1\right) \bigg\} \\ &&\times \left\Vert u-v\right\Vert . \end{eqnarray} $ |
Taking the supermum over all $ t\in [0, T] $ and using the inequality (3.14) yield
$ \begin{eqnarray} \left\Vert \Phi _{1}u-\Phi _{1}v\right\Vert &\leq &\left\Vert B_{1}u-B_{1}v\right\Vert +\left\Vert Cu-Cv\right\Vert \\ &\leq &\Delta \Vert u-v\Vert . \end{eqnarray} $ |
From the condition (4.3), we conclude that the operator $ \Phi _{1} $ is a contraction mapping. Hence, the problem (4.1) has a unique solution on $ [0, T]. $ The proof is completed.
Now, for $ \sigma = 1 $, using Remark 19 above, we can easily come by the following result.
Corollary 24. Let $ \sigma = 1 $ (i.e., $ _{p}^{\mathcal{C}}\mathfrak{D} _{0^{+}}^{\beta, \sigma } = $ $ ^{C}D_{0^{+}}^{\beta } $) and assume that $ (H_{2}) $ and $ (H_{4}) $ are satisfied. Then, the problem describe by Eq (4.1) has a unique solution on $ [0, T] $ if
$ \begin{equation} \Delta _{1} < 1, \end{equation} $ | (4.6) |
where
$ \begin{equation} \Delta _{1} = \frac{2L_{g}T^{\beta }}{\Gamma (\beta +1)}+\sum\limits_{i = 1}^{i = m}\frac{ 2L_{h_{i}}T^{\alpha _{i}}}{\Gamma (\alpha _{i}+1)}. \end{equation} $ | (4.7) |
We begin introducing the concept of Ulam-type stability for the problem described by (4.1).
Definition 25. The solution of the problem described by (4.1) is said to be Ulam-Hyers stable if there exists a real constant $ K_{g} > 0 $ such that for given $ \varepsilon > 0 $ and for each solution $ v\in {C([0, T], \mathbb{R})} $ of the inequality
$ \begin{equation} \left\vert \text{ }_{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma } \bigg(u\left( t\right) -\sum\limits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))\bigg)-g\left( t, u\left( t\right) \right) \right\vert \leq \varepsilon , \end{equation} $ | (5.1) |
there exists a solution $ u\in {C([0, T], \mathbb{R})} $ of problem described by Eq (4.1) with
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert < K_{g}\varepsilon , \quad \text{for}\quad t\in {[0, T]}. \end{equation*} $ |
Definition 26. The solution of the problem described by Eq (4.1) is said to be generalized Ulam-Hyers stable if there exists a function $ \phi _{g}\in {C(\mathbb{R}^{+}, \mathbb{R} ^{+})} $ with $ \phi _{g}(0) = 0 $ such that, for any given $ \varepsilon > 0 $ and for each solution $ v\in {C([0, T], \mathbb{R})} $ of the inequality (5.1), there exists a solution $ u\in {C([0, T], \mathbb{R})} $ of problem described by (4.1) with
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert < \phi _{g}(\varepsilon ), \quad \ {for} \ \quad t\in {[0, T]}. \end{equation*} $ |
The stability mentioned in Definition 25 can be generalized if the constant $ \varepsilon $ is replaced by a certain type of functions. Such stability is called the Ulam-Hyers-Rassias stability.
Definition 27. The solution of the problem described by (4.1) is said to be Ulam-Hyers-Rassias stable with respect to $ \phi _{g}\in {C}([0, T], \mathbb{R}^{+}) $ if there exists $ K_{g, \phi } > 0 $ such that, for any given $ \varepsilon > 0 $ and for each solution $ v\in {C}([0, T], \mathbb{R}) $ of the inequality
$ \begin{equation} \left\vert \text{ }_{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma } \bigg(u\left( t\right) -\sum\limits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))\bigg)-g\left( t, u\left( t\right) \right) \right\vert \leq \varepsilon \phi _{g}(t){, } \end{equation} $ | (5.2) |
there exists a solution $ u\in {C([0, T], \mathbb{R})} $ of the problem described by (4.1) with
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert < K_{g, \phi }\varepsilon \phi _{g}(t), \quad \ {for}\quad t\in {[0, T]}. \end{equation*} $ |
Definition 28. The solution of the problem described by (4.1) is a generalized Ulam-Hyers-Rassias stable with respect to $ \phi _{g}\in {C}([0, T], \mathbb{R}^{+}) $ if there exists $ K_{g, \phi } > 0 $ such that, for any given $ \varepsilon > 0 $ and for each solution $ v\in {C([0, T], \mathbb{R})} $ of inequality
$ \begin{equation} \left\vert \text{ }_{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta , \sigma } \bigg(u\left( t\right) -\sum\limits_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t))\bigg)-g\left( t, u\left( t\right) \right) \right\vert \leq \phi _{g}(t), \end{equation} $ | (5.3) |
there exists a solution $ u\in {C([0, T], \mathbb{R})} $ of problem (4.1) with
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert < K_{g, \phi }\phi _{g}(t), \quad \ {for} \ \quad t\in {[0, T]}. \end{equation*} $ |
Remark 29. A function $ v\in {C([0, T], \mathbb{R})} $ is a solution of the inequality (5.2) if and only if there exists a small perturbation $ \Xi \in C([0, T], \mathbb{R}) $ (dependent on $ v $) such that
i) $ \left\vert \Xi (t)\right\vert \leq \varepsilon \phi _{\Xi }(t), $ $ t\in { [0, T], } $
ii) $ _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\beta, \sigma }\bigg(u\left(t\right) -\sum_{i = 1}^{i = m}J_{0^{+}}^{\alpha _{i}, \sigma }h_{i}(t, u(t)) \bigg) = g\left(t, u\left(t\right) \right) +\Xi (t), $ $ t\in {[0, T]}. $
Lemma 30. If $ v\in {C}([0, T], \mathbb{R}) $ represents a solution of inequality (5.2), then $ v $ is a solution of the following integral inequality
$ \begin{equation} \left\vert v(t)-\Phi _{1}v(t)\right\vert \leq \varepsilon \Pi _{\phi _{\Xi }}(t) \end{equation} $ | (5.4) |
where
$ \begin{equation} \Pi _{\phi _{\Xi }}(t) = \frac{t}{T}e^{\frac{1-\sigma }{\sigma }(T-t)}( \mathcal{J}_{0^{+}}^{\alpha , \sigma }\phi _{\Xi })(T)+(\mathcal{J} _{0^{+}}^{\alpha , \sigma }\phi _{\Xi })(t) \end{equation} $ | (5.5) |
Proof. From Remark 29, we get
$ \begin{eqnarray} v(t) & = &\frac{u_{0}}{f(0, u_{0})}\left( 1-\frac{t}{T}\right) e^{\frac{\sigma -1}{\sigma }t}+\frac{u_{T}\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{ f(T, u_{T})e^{\frac{\sigma -1}{\sigma }T}} \\ &&-\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{f(T, u_{T})e^{\frac{ \sigma -1}{\sigma }T}}\sum\limits_{i = 1}^{i = m}\frac{\int_{0}^{T}\left( T-\rho \right) ^{\alpha _{i}-1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}h_{i}(\rho , v\left( \rho \right) )d\rho }{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) } \\ &&-\frac{\frac{t}{T}e^{\frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{\frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(T-\rho )}\left[ g\left( \rho , v\left( \rho \right) \right) +\Xi (\rho )\right] d\rho \\ &&+\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\left[ g\left( \rho , v\left( \rho \right) \right) +\Xi (\rho )\right] d\rho \\ &&+\sum\limits_{i = 1}^{i = m}\frac{1}{\sigma ^{\alpha _{i}}\Gamma \left( \alpha _{i}\right) }\int_{0}^{t}\left( t-\rho \right) ^{\alpha _{i}-1}e^{\frac{ \sigma -1}{\sigma }(t-\rho )}h_{i}(\rho , u\left( \rho \right) )d\rho , \end{eqnarray} $ |
this yields that
$ \begin{eqnarray} \left\vert v(t)-\Phi _{1}v(t)\right\vert &\leq &\bigg|-\frac{\frac{t}{T}e^{ \frac{\sigma -1}{\sigma }t}}{\sigma ^{\beta }\Gamma \left( \beta \right) e^{ \frac{\sigma -1}{\sigma }T}}\int_{0}^{T}\left( T-\rho \right) ^{\beta -1}e^{ \frac{\sigma -1}{\sigma }(T-\rho )}\Xi (\rho )d\rho \\ &&+\frac{1}{\sigma ^{\beta }\Gamma \left( \beta \right) }\int_{0}^{t}\left( t-\rho \right) ^{\beta -1}e^{\frac{\sigma -1}{\sigma }(t-\rho )}\Xi (\rho )d\rho \bigg| \\ & = &\frac{t}{T}e^{\frac{1-\sigma }{\sigma }(T-t)}\text{ }\left\vert (\mathcal{ \ J}_{0^{+}}^{\alpha , \sigma }\Xi )(T)+(\mathcal{J}_{0^{+}}^{\alpha , \sigma }\Xi )(t)\right\vert \\ &\leq &\frac{t}{T}e^{\frac{1-\sigma }{\sigma }(T-t)}\varepsilon (\mathcal{J} _{0^{+}}^{\alpha , \sigma }\phi _{\Xi })(T)+\varepsilon (\mathcal{J} _{0^{+}}^{\alpha , \sigma }\phi _{\Xi })(t) \\ & = &\varepsilon \Pi _{\phi _{\Xi } }(t) , \end{eqnarray} $ |
which leads to the inequality in (5.4).
In the following, we present the theorem related the Ulam-Hyers-Rassias stability of the solution of the problem described by Eq (4.1).
Theorem 31. Let $ \sigma \in (0, 1). $ Assume that $ (H_{1}), $ $ (H_{2}) $ and$ \ (H_{4}) $ are satisfied with
$ \begin{equation*} \Delta < 1. \end{equation*} $ |
Then, the solution to the problem described by Eq (4.1) is both Ulam-Hyers-Rassias stable and generalized Ulam-Hyers Rassias stable on $ [0, T]. $
Proof. From Lemma 30 we get, for $ t\in [0, T] $
$ \begin{eqnarray} \left\vert v(t)-u(t)\right\vert &\leq &\left\vert v(t)-\Phi _{1}v(t)\right\vert +\left\vert \Phi _{1}v(t)-u(t)\right\vert \\ & = &\left\vert v(t)-\Phi _{1}v(t)\right\vert +\left\vert \Phi _{1}v(t)-\Phi _{1}u(t)\right\vert \\ &\leq &\varepsilon \Pi _{\phi _{\Xi }}(t)+\Delta \left\vert v(t)-u(t)\right\vert . \end{eqnarray} $ |
This yields that
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert \leq \dfrac{\varepsilon \Pi _{\phi _{\Xi }}(t)}{1-\Delta }. \end{equation*} $ |
By setting $ K_{g, \phi } = \frac{1}{1-\Delta } $, we end up with
$ \begin{equation*} \left\vert v(t)-u(t)\right\vert \leq K_{g, \phi }\varepsilon \Pi _{\phi _{\Xi }}(t). \end{equation*} $ |
Hence, the solution of problem described by Eq (4.1) is Ulam-Hyers-Rassias stable with respect to $ \Pi _{\phi _{\Xi }}. $ Moreover, if we set $ \phi _{g}(t) = \varepsilon \Pi _{\phi _{\Xi }}(t) $ with $ \phi _{g}(0) = 0, $ then the same solution is generalized Ulam-Hyers-Rassias stable. The proof is completed.
Corollary 32. Let $ \sigma = 1 $ (i.e., $ _{p}^{\mathcal{C}}\mathfrak{D} _{0^{+}}^{\beta, \sigma } = $ $ ^{C}D_{0^{+}}^{\beta } $) and assume that $ (H_{2}) $ and $ (H_{4}) $ are satisfied with $ \Delta _{1} < 1 $ where $ \Delta _{1} $ is given by (4.7). Then, the solution of the problem described by Eq (4.1) is both Ulam-Hyers-Rassias stable and generalized Ulam-Hyers Rassias stable on $ [0, T]. $
Remark 33. Let $ \sigma \in (0, 1]. $ By using Theorem 31 and Corollary 32, we can conclude that:
1) If $ \Pi _{\phi _{\Xi }}(t) = 1, $ then the solution of the problem described by (4.1) is Ulam-Hyers stable.
2) If we set $ \phi _{g}(\varepsilon) = K_{g, \phi }\varepsilon $ with $ \phi _{g}(0) = 0 $, then the solution of the problem described by (4.1) is generalized Ulam-Hyers stable.
Consider the following hybrid proportional fractional integro-differential equation$ : $
$ \begin{equation} \left\{ \begin{array}{l} _{p}^{\mathcal{C}}\mathfrak{D}_{0^{+}}^{\frac{3}{2}, \frac{3}{4}}\left( \frac{ u\left( t\right) -\sum\nolimits_{i = 1}^{i = 4}J_{0^{+}}^{\frac{2i-1}{4}, \frac{3}{4}} \frac{\cos \left\vert u\left( t\right) \right\vert }{t+10i}}{f(t, u(t))} \right) = \frac{\epsilon (1-t)\sin \left\vert u(t)\right\vert }{1+\left\vert u(t)\right\vert }, t\in [0, 1], \\ u(0) = \frac{1}{10}, \ u(1) = \frac{1}{20}, \end{array} \right. \end{equation} $ | (6.1) |
where the real constant $ \epsilon $ and the function $ f(t, u(t)) $ will be fixed later.
Here $ T = 1, \beta = \frac{3}{2}, \sigma = \frac{3}{4}, \alpha _{i} = \frac{2i-1}{4}, h_{i}(t, u(t)) = \frac{\cos \left\vert u\left(t\right) \right\vert }{t+100i}, (i = 1, ..., 4), g(t, u(t)) = \frac{ \epsilon (1-t)\sin \left\vert u(t)\right\vert }{1+\left\vert u(t)\right\vert }, u_{0} = \frac{1}{10} $ and $ u_{1} = \frac{1}{20}. $
We can show that
$ \begin{equation*} \left\vert h_{i}(t, \lambda _{1})-h_{i}(t, \lambda _{2})\right\vert \leq \frac{ 1}{10i}\left\vert \lambda _{1}-\lambda _{2}\right\vert , \ \ i = 1, ..., 4, \end{equation*} $ |
$ \begin{equation*} \left\vert h_{i}(t, 0)\right\vert \leq \frac{1}{10i}, \ \ i = 1, ..., 4, \end{equation*} $ |
$ \begin{equation*} \left\vert g(t, u(t))\right\vert \leq \left\vert \epsilon \right\vert (1-t) \frac{\left\Vert u\right\Vert }{1+\left\Vert u\right\Vert }. \end{equation*} $ |
Also, for all $ (t, \lambda)\in [0, 1]\times \mathbb{R}, $ we have
$ \begin{eqnarray} \left\vert \partial _{\lambda }g(t, \lambda )\right\vert & = &\left\vert \epsilon \right\vert (1-t)\left\vert \frac{(1+\left\vert \lambda \right\vert )\cos \left\vert \lambda \right\vert -\sin \left\vert \lambda \right\vert }{ (1+\left\vert \lambda \right\vert )^{2}}\right\vert \\ &\leq &2\left\vert \epsilon \right\vert , \end{eqnarray} $ |
It follows that $ L_{h_{i}} = \frac{1}{10i} = \left\Vert \theta _{i}\right\Vert, (i = 1, ..., 4), L_{g} = 2\left\vert \epsilon \right\vert, \left\Vert \widetilde{g} \right\Vert = \left\vert \epsilon \right\vert $ and $ \psi (\left\Vert u\right\Vert) = \frac{\left\Vert u\right\Vert }{1+\left\Vert u\right\Vert }. $
Using Matlab program, we find
$ \begin{equation*} P(\beta , \frac{1-\sigma }{\sigma }T) = 0.118985157486215, \end{equation*} $ |
$ \begin{equation*} P(\alpha _{1}, \frac{1-\sigma }{\sigma }T) = 0.787212464733354, \end{equation*} $ |
$ \begin{equation*} P(\alpha _{2}, \frac{1-\sigma }{\sigma }T) = 0.415815178677325, \end{equation*} $ |
$ \begin{equation*} P(\alpha _{3}, \frac{1-\sigma }{\sigma }T) = 0.186545461450941, \end{equation*} $ |
$ \begin{equation*} P(\alpha _{4}, \frac{1-\sigma }{\sigma }T) = 0.073797013512809. \end{equation*} $ |
For illustrating Theorem 18, we take $ f(t, u(t)) = \frac{1}{10}\sqrt{ 1+\left\vert u(t)\right\vert ^{2}} $ and $ \epsilon = 1. $
We can show that
$ \begin{equation*} \left\vert f(t, \lambda _{1})-f(t, \lambda _{2})\right\vert \leq \frac{1}{10} \left\vert \lambda _{1}-\lambda _{2}\right\vert \text{ and }\left\vert f(t, 0)\right\vert = \frac{1}{10}, \end{equation*} $ |
i.e., $ L_{f} = \frac{1}{10} $ and $ \left\Vert \zeta \right\Vert = \frac{1}{10}. $
We see that the condition (3.7) and (3.8) is followed with a real number $ R\in $ $ [0.929854125359180, 5.322899777680700]. $
Therefore, all conditions in Theorem 18 are satisfied. We conclude that the problem (6.1) has at least one solution on $ [0, 1]. $
Remark 34. If $ R = 0.929854125359180, $ we obtain
$ \begin{equation*} \varrho = 2.555690432688444, \end{equation*} $ |
$ \begin{equation*} \frac{\varrho \left\Vert \zeta \right\Vert +\sum\nolimits_{i = 1}^{i = m}\frac{\left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}}{1-\varrho L_{f}-\sum\nolimits_{i = 1}^{i = m}\frac{ L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}} = 0.929854125359179\leq R, \end{equation*} $ |
and
$ \begin{equation*} \varrho L_{f}+\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{ \sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} = 0.481751141209855 < 1, \end{equation*} $ |
Also, if $ R = 5.322899777680700, $ we obtain
$ \begin{equation*} \varrho = 6.153559316370539, \end{equation*} $ |
$ \begin{equation*} \frac{\varrho \left\Vert \zeta \right\Vert +\sum\nolimits_{i = 1}^{i = m}\frac{\left\Vert \theta _{i}\right\Vert P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}}{1-\varrho L_{f}-\sum\nolimits_{i = 1}^{i = m}\frac{ L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{\sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}}} = 5.322899777680694\leq R, \end{equation*} $ |
and
$ \begin{equation*} \varrho L_{f}+\sum\limits_{i = 1}^{i = m}\frac{L_{h_{i}}P(\alpha _{i}, \frac{1-\sigma }{ \sigma }T)}{\left( 1-\sigma \right) ^{\alpha _{i}}} = 0.841538029578065 < 1. \end{equation*} $ |
Note that all conditions in Theorem 18 are satisfied. We conclude that the problem (6.1) has at least one solution on $ [0, 1]. $
For illustrating Theorems 23 and 31, we take $ f(t, \lambda) = 1 $ for all $ (t, \lambda)\in $ $ [0, 1]\times < /p > < p > \mathbb{R} < /p > < p > $ and $ \epsilon = \frac{1}{11\pi }. $
Using the above data, we find:
$ \begin{equation*} \Delta = 0.978416714361031 < 1. \end{equation*} $ |
In virtue of Theorem 23, problem (6.1) has a unique solution. Furthermore, we can compute
$ \begin{equation*} K_{g, \phi } = \frac{1}{1-\Delta } = 46.332148715785415 > 0. \end{equation*} $ |
Thus, by use of Theorem 31, problem (6.1) is Ulam-Hyers-Rassias stable, and consequently generalized Ulam-Hyers-Rassias stable.
In this work, we inspected the existence of solutions to a certain class of hybrid fractional intego-differential equations in the casement of generalized proportional fractional hybrid integro-differential equations supplemented with Dirichlet boundary conditions. The existence of at least one solution was shown with the assistance of the hybrid fixed point theorem for a product of three operators. In addition, for some special cases, the uniqueness of the solutions for the considered class of equations, and their stability in the sense of Ulam, were discussed.
We believe that the results obtained will be very important for the researchers working on the qualitative aspects of the boundary value problems in the frame work of the generalized fractional derivatives mentioned in the article. This is due to the fact that as far as we know this is the first work in which the lower regularized incomplete gamma function is used to discuss some qualitative aspects of solutions to fractional differential equations.
The authors declare there is no conflict of interest.
[1] |
J. T. Machado, V. Kiryakova, F. Mainardi, Recent history of fractional calculus, Commun. Nonlinear Sci. Numer. Simul., 16 (2011), 1140–1153. https://doi.org/10.1016/j.cnsns.2010.05.027 doi: 10.1016/j.cnsns.2010.05.027
![]() |
[2] | R. Hilfer, Threefold introduction to fractional derivatives, In: Anomalous transport, foundations and applications, Wiley, 2008. https://doi.org/10.1002/9783527622979.ch2 |
[3] | K. S. Miller, B. Ross, An introduction to the fractional calculus and fractional differential equations, New York: Wiley, 1993. |
[4] | J. Sabatier, O. P. Agrawal, J. A. T. Machado, Advances in fractional calculus: theoretical developments and applications in physics and engineering, Dordrecht: Springer, 2007. https://doi.org/10.1007/978-1-4020-6042-7 |
[5] | F. Maindari, Fractional calculus: some basic problems in continuum and statistical mechanics, In: Fractals and fractional caluculas in continuum mechanics, Vienna: Springer, 1997,291–348. https://doi.org/10.1007/978-3-7091-2664-6_7 |
[6] |
J. A. T. Machado, And I say to myself: "What a fractional world!", Fract. Calc. Appl. Anal., 14 (2011), 635. https://doi.org/10.2478/s13540-011-0037-1 doi: 10.2478/s13540-011-0037-1
![]() |
[7] |
J. A. T. Machado, F. Mainardi, V. Kiryakova, Fractional calculus: quo vadimus? (Where are we going?), Fract. Calc. Appl. Anal., 18 (2015), 495–526. https://doi.org/10.1515/fca-2015-0031 doi: 10.1515/fca-2015-0031
![]() |
[8] |
H. G. Sun, Y. Zhang, D. Baleanu, W. Chen, Y. Q. Chen, A new collection of real world applications of fractional calculus in science and engineering, Commun. Nonlinear Sci. Numer. Simul., 64 (2018), 213–231. https://doi.org/10.1016/j.cnsns.2018.04.019 doi: 10.1016/j.cnsns.2018.04.019
![]() |
[9] |
W. Qiu, D. Xu, J. Guo, Numerical solution of the fourth-order partial integro-differential equation with multi-term kernels by the Sinc-collocation method based on the double exponential transformation, Appl. Math. Comput., 392 (2021), 125693. https://doi.org/10.1016/j.amc.2020.125693 doi: 10.1016/j.amc.2020.125693
![]() |
[10] |
D. Xu, W. Qiu, J. Guo, A compact finite difference scheme for the fourth-order time-fractional integro-differential equation with a weakly singular kernel, Numer. Meth. Part Differ. Equ., 36 (2020), 439–458. https://doi.org/10.1002/num.22436 doi: 10.1002/num.22436
![]() |
[11] | K. B. Ali, A. Ghanmi, K. Kefi, Existence of solutions for fractional differential equations with Dirichlet boundary conditions, Electron. J. Diff. Equ., 2016 (2016), 116. |
[12] |
R. A. C. Ferreira, Fractional de la Vallée Poussin inequalities, Math. Inequal. Appl., 22 (2019), 917–930. https://doi.org/10.7153/mia-2019-22-62 doi: 10.7153/mia-2019-22-62
![]() |
[13] | R. A. C. Ferreira, Existence and uniqueness of solutions for two-point fractional boundary value problems, Electron. J. Diff. Equ., 2016 (2016), 202. |
[14] | Z. Laadjal, N. Adjeroud, Sharp estimates for the unique solution of the Hadamard-type two-point fractional boundary value problems, Appl. Math. E-Notes, 2021 (2021), 275–281. |
[15] |
R. P. Agarwal, R. Luca, Positive solutions for a semipositone singular Riemann-Liouville fractional differential problem, Int. J. Nonlinear Sci. Numer. Simul., 20 (2019), 823–831. https://doi.org/10.1515/ijnsns-2018-0376 doi: 10.1515/ijnsns-2018-0376
![]() |
[16] |
F. Jiao, Y. Zhou, Existence of solution for a class of fractional boundary value problems via critical point theory, Comput. Math. Appl., 62 (2011), 1181–1199. https://doi.org/10.1016/j.camwa.2011.03.086 doi: 10.1016/j.camwa.2011.03.086
![]() |
[17] | S. T. M. Thabet, M. B. Dhakne, On boundary value problems of higher order abstract fractional integro-differential equations, Int. J. Nonlinear Anal. Appl., 7 (2016), 165–184. |
[18] |
A. Anguraj, P. Karthikeyan, J. J. Trujillo, Existence of solutions to fractional mixed integrodifferential equations with nonlocal initial condition, Adv. Differ. Equ., 2011 (2011), 690653. https://doi.org/10.1155/2011/690653 doi: 10.1155/2011/690653
![]() |
[19] |
B. Ahmad, J. J. Nieto, Existence results for nonlinear boundary value problems of fractional integrodifferential equations with integral boundary conditions, Bound. Value Probl., 2009 (2009), 708576. https://doi.org/10.1155/2009/708576 doi: 10.1155/2009/708576
![]() |
[20] | H. L. Tidke, Existence of global solutions to nonlinear mixed Volterra-Fredholm integrodifferential equations with nonlocal conditions, Electron. J. Diff. Equ., 2009 (2009), 55. |
[21] |
B. Ahmad, J. J. Nieto, Boundary value problems for a class of sequential integrodifferential equations of fractional order, J. Funct. Space., 2013 (2013), 149659. https://doi.org/10.1155/2013/149659 doi: 10.1155/2013/149659
![]() |
[22] |
D. H. Hyers, On the stability of the linear functional equation, Proc. Natl. Acad. Sci. USA, 27 (1941), 222–224. https://doi.org/10.1073/pnas.27.4.222 doi: 10.1073/pnas.27.4.222
![]() |
[23] |
A. Ali, F. Rabiei, K. Shah, On Ulam's type stability for a class of impulsive fractional differential equations with nonlinear integral boundary conditions, J. Nonlinear Sci. Appl., 10 (2017), 4760–4775. https://doi.org/10.22436/jnsa.010.09.19 doi: 10.22436/jnsa.010.09.19
![]() |
[24] |
K. Shah, A. Ali, S. Bushnaq, Hyers-Ulam stability analysis to implicit Cauchy problem of fractional differential equations with impulsive conditions, Math. Method. Appl. Sci., 41 (2018), 8329–8343. https://doi.org/10.1002/mma.5292 doi: 10.1002/mma.5292
![]() |
[25] |
A. Ali, K. Shah, D. Baleanu, Ulam stability results to a class of nonlinear implicit boundary value problems of impulsive fractional differential equations, Adv. Differ. Equ., 2019 (2019), 5. https://doi.org/10.1186/s13662-018-1940-0 doi: 10.1186/s13662-018-1940-0
![]() |
[26] |
Asma, A. Ali, K. Shah, F. Jarad, Ulam-Hyers stability analysis to a class of nonlinear implicit impulsive fractional differential equations with three point boundary conditions, Adv. Differ. Equ., 2019 (2019), 7. https://doi.org/10.1186/s13662-018-1943-x doi: 10.1186/s13662-018-1943-x
![]() |
[27] |
A. Ali, K. Shah, F. Jarad, V. Gupta, T. Abdeljawad, Existence and stability analysis to a coupled system of implicit type impulsive boundary value problems of fractional-order differential equations, Adv. Differ. Equ., 2019 (2019), 101. https://doi.org/10.1186/s13662-019-2047-y doi: 10.1186/s13662-019-2047-y
![]() |
[28] |
B. C. Dhage, V. Lakshmikantham, Basic results on hybrid differential equations, Nonlinear Anal. Hybrid Syst., 4 (2010), 414–424. https://doi.org/10.1016/j.nahs.2009.10.005 doi: 10.1016/j.nahs.2009.10.005
![]() |
[29] |
M. A. E. Herzallah, D. Baleanu, On fractional order hybrid differential equations, Abstr. Appl. Anal., 2014 (2014), 389386. https://doi.org/10.1155/2014/389386 doi: 10.1155/2014/389386
![]() |
[30] |
B. Ahmad, S. K. Ntouyas, J. Tariboon, On hybrid Caputo fractional integro-differential inclusions with nonlocal conditions, J. Nonlinear Sci. Appl., 9 (2016), 4235–4246. https://doi.org/10.22436/jnsa.009.06.65 doi: 10.22436/jnsa.009.06.65
![]() |
[31] | B. C. Dhage, G. T. Khrpe, A. Y. Shete, J. N. Salunke, Existence and approximate solutions for nonlinear hybrid fractional integrodifferential equations, Int. J. Anal. Appl., 11 (2016), 157–167. |
[32] |
Z. Laadjal, T. Abdeljawad, F. Jarad, On existence-uniqueness results for proportional fractional differential equations and incomplete gamma functions, Adv. Differ. Equ., 2020 (2020), 641. https://doi.org/10.1186/s13662-020-03043-8 doi: 10.1186/s13662-020-03043-8
![]() |
[33] | A. A. Kilbas, H. M. Srivastava, J. J. Trujillo, Theory and applications of fractional differential equations, Amsterdam: Elsevier, 2006. |
[34] |
F. Jarad, T. Abdeljawad, J. Alzabut, Generalized fractional derivatives generated by a class of local proportional derivatives, Eur. Phys. J. Spec. Top., 226 (2017), 3457–3471. https://doi.org/10.1140/epjst/e2018-00021-7 doi: 10.1140/epjst/e2018-00021-7
![]() |
[35] | A. Erdélyi, W. Magnus, F. Oberhettinger, F. G. Tricomi, Higher transcendental functions, Vol. II, New York, Toronto, London: McGraw-Hill Book Company, 1953. |
[36] |
A. Gil, J. Segura, N. M. Temme, Efficient and accurate algorithms for the computation and inversion of the incomplete gamma function ratios, SIAM J. Sci. Comput., 34 (2012), A2965–A2981. https://doi.org/10.1137/120872553 doi: 10.1137/120872553
![]() |
[37] |
Z. Laadjal, F. Jarad, On a Langevin equation involving Caputo fractional proportional derivatives with respect to another function, AIMS Mathematics, 7 (2022), 1273–1292. https://doi.org/10.3934/math.2022075 doi: 10.3934/math.2022075
![]() |
[38] | B. C. Dhage, A fixed point theorem in Banach ilgebras Involving three operators with applications, Kyungpook Math. J., 44 (2004), 145–155. |
1. | Bashir Ahmad, Manal Alnahdi, Sotiris K. Ntouyas, Existence Results for a Differential Equation Involving the Right Caputo Fractional Derivative and Mixed Nonlinearities with Nonlocal Closed Boundary Conditions, 2023, 7, 2504-3110, 129, 10.3390/fractalfract7020129 | |
2. | Ravi P. Agarwal, Bashir Ahmad, Hana Al-Hutami, Ahmed Alsaedi, Existence results for nonlinear multi-term impulsive fractional $ q $-integro-difference equations with nonlocal boundary conditions, 2023, 8, 2473-6988, 19313, 10.3934/math.2023985 | |
3. | Zahra Eidinejad, Reza Saadati, Javad Vahidi, Chenkuan Li, Tofigh Allahviranloo, The existence of a unique solution and stability results with numerical solutions for the fractional hybrid integro-differential equations with Dirichlet boundary conditions, 2024, 2024, 1687-2770, 10.1186/s13661-024-01928-1 | |
4. | AHMED ALSAEDI, HANA AL-HUTAMI, BASHIR AHMAD, INVESTIGATION OF A NONLINEAR MULTI-TERM IMPULSIVE ANTI-PERIODIC BOUNDARY VALUE PROBLEM OF FRACTIONAL q-INTEGRO-DIFFERENCE EQUATIONS, 2023, 31, 0218-348X, 10.1142/S0218348X23401916 | |
5. | Ahmed Alsaedi, Bashir Ahmad, Hana Al-Hutami, Nonlinear Multi-term Impulsive Fractional q-Difference Equations with Closed Boundary Conditions, 2024, 23, 1575-5460, 10.1007/s12346-023-00934-5 | |
6. | Ahmed Alsaedi, Manal Alnahdi, Bashir Ahmad, Sotiris K. Ntouyas, On a nonlinear coupled Caputo-type fractional differential system with coupled closed boundary conditions, 2023, 8, 2473-6988, 17981, 10.3934/math.2023914 | |
7. | Madeaha Alghanmi, Ravi P. Agarwal, Bashir Ahmad, Existence of Solutions for a Coupled System of Nonlinear Implicit Differential Equations Involving $$\varrho $$-Fractional Derivative with Anti Periodic Boundary Conditions, 2024, 23, 1575-5460, 10.1007/s12346-023-00861-5 | |
8. | Boshra Alharbi, Ahmed Alsaedi, Ravi P. Agarwal, Bashir Ahmad, Existence results for a nonlocal q-integro multipoint boundary value problem involving a fractional q-difference equation with dual hybrid terms, 2024, 32, 1844-0835, 5, 10.2478/auom-2024-0026 |