Abstract
The aim of this manuscript is two-fold. First, we analyze a linear fractional differential equation for the existence, uniqueness and Ulam–Hyers stabilities and then we consider a nonlinear fractional differential equation for the said analysis. For the linear problem, an important tool that is, modified Laplace transform is considered. By employing the \(\rho \)-Laplace transform method, we verify that the linear fractional matrix-valued differential equation is Ulam–Hyers stable. For the nonlinear fractional integro-differential equation the arguments for the existence and uniqueness of solutions are on account of Schaefer’s fixed point approach and Banach contraction principle. Through extended Gronwall’s inequality, we achieve the stability results. At the end, with the help of an example, we demonstrate our main theoretical work.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
In the last few decades, much attention has been paid to Fractional Calculus (FC) due to its wide range applications into real world phenomena. Different attempts have been made in order to present these phenomena in a superior way and to explore new fractional operators with different kernels such as Riemann–Liouville, Caputo, Hadamard, and generalized Liouville–Caputo [3, 7, 22, 29]. In [8], a new and practical fractional derivative operator has been introduced called the Atangana–Baleanu fractional derivative.
Integral inequalities has a major contribution in flourishing the analysis of differential and integral equations. Among the well known inequalities, Gronwall inequality [10] is a most popular one, because of its applications in many areas of mathematics. A generalized Gronwall inequality on a Riemann–Liouville fractional differential equations was considered in [30]. While in [1], the Gronwall inequality was obtained for q-fractional operators. Liu et al. [22] presented the extended Gronwall inequality for the analysis of Liouville–Caputo fractional differential equations.
In the literature of FC, there exist many stability perceptions, for instance: Mittag–Leffler stability, asymptotic stability, global asymptotic uniform stability, synchronization problems, stabilization problems and fractional input stability [20, 26]. But the most simple and easy to use when dealing with models which describe real world phenomena is the Ulam stability, mentioned by the Ulam [27] in 1940, then splendidly solved by Hyers [15] in 1941 and called it Ulam–Hyers (UH) stability. Subsequently, many renown mathematicians explored the above said stability brilliantly; for details, one can see [4,5,6, 9, 11,12,13, 24, 31,32,33]. Razaei et al. [25] studied the UH stability of linear differential equations. Recently in [28], the authors presented the UH stability of two types of linear differential equations.
In most of our analysis, when dealing with differential equations including physical models, there are some strong and interesting tools found in the literature like, Riemann–Liouville, Caputo and Riez fractional derivatives, integral transforms including Laplace, Mellin and Fourier for finding the exact solution. Recently, a new generalized fractional derivatives and integrals presented by Katugampola [18, 19], which consolidates the Riemann–Liouville and the Hadamard derivatives and integrals respectively.
Liu et al. [22] investigated the UH stability of linear differential equations with the help of \(\rho \)-Laplace transform. After that, they studied the following Cauchy problem of nonlinear equations
where \(\rho >0\), \(0<\gamma <1\).
Inspired by the work mentioned above [22], our first task is to investigate the following fractional differential equations for the said analysis
where \({^{c}D_{t_0}^{\gamma ,\rho }}\) denotes the left generalized \(\gamma \) order Liouville–Caputo fractional derivative defined componentwise. A is n-th order matrix over the real field \({\mathbb {R}}\), q(t) is n-dimensional locally integrable column vector function on the closed interval \([t_0,T]\) and its entries admits the \(\rho \)-Laplace transformation, \({\mathcal {W}}(t)=(w_{1}(t),w(t),\dots ,w_{n}(t))^{T}\) is an unknown vector function, while P is a specified vector.
Then, we consider the following integro-differential equation with generalized Liouville–Caputo fractional derivative of the form
where \(\rho >0\), \(\gamma \in (0,1]\), \(^{c}D_{t_{0+}}^{\gamma ,\rho }\alpha (t)\) denotes the left generalized \(\gamma \) order Liouville–Caputo fractional derivative for \(\alpha \) with the parameter \(\rho \), \(l:\Delta \times {\mathbb {R}} \rightarrow {\mathbb {R}}\). For the sake of simplicity, we take
The manuscript is organized as follows: In section “Preliminaries”, we recall some important definitions, lemmas and basic properties related to \(\rho \)-Laplace transform. In section “Stability Analysis of (1)”, we concentrate on the stability of problem (1). Section “Existence Result of Nonlinear Model (2)” is devoted to the existence and uniqueness of problem (2). Section “Stability Analysis for Nonlinear Model” is managed to discuss the stability analysis of problem (2). In section “Example”, we mention the applications of our obtained results. Last Section of this article is dedicated to the conclusion.
Preliminaries
In this section, we discuss some imperative definitions of generalized fractional derivatives and integrals. To prove our main results, we state some basic definitions, lemmas and properties of the \(\rho \)-Laplace transform.
Assume that the space of all continuous functions from J into \({\mathbb {R}}\) is denoted by \(C(J,{\mathbb {R}})\) with norm:
Definition 2.1
[19] Let \(\theta :[a,\infty )\rightarrow {\mathbb {R}}\), then the generalized left-sided integral is defined as
where \(\gamma \in (0,1)\), \(\rho >0\).
Similarly, the generalized right-sided integral is defined as
Definition 2.2
[18] Suppose that \(0<\gamma <1\), \(\rho >0\). The generalized Liouville–Caputo derivative of the function \(\theta :[0,+\infty )\rightarrow {\mathbb {R}}\) is expressed in the form
Definition 2.3
[23] The standard Mittag–Leffler function for \(a \in {\mathbb {C}}\), the set of all complex numbers, defined by the series
where \(\Gamma (\cdot )\) is the gamma function.
The two parameter Mittag–Leffler function for \(a,b \in {\mathbb {C}}\) is given by
where \({\mathbb {C}}\) denotes the set of all complex numbers. The series is convergent when a and b are strictly positive. It is abbreviated as \({\mathbb {E}}_{a}(z)={\mathbb {E}}_{a,1}(z)\), when \(b=1\). We acquired the classical exponential function, when \(a=b=1\).
Definition 2.4
The Mittag–Leffler matrix \({\mathbb {E}}_{a}(A)\) is defined as;
where \(a>0\) and \(A_{n} \in M_{n}\).
Theorem 2.5
[16] Let \(\rho >0\), \(\gamma \in (0,1)\). The \(\rho \)-Laplace transform of the function z of the generalized fractional derivative in the Liouville–Caputo sense can be defined as
The \(\rho \)-Laplace transform of the function z is defined as;
Then
where, \({\mathcal {L}}_{\rho }\{z\}\) represents the usual Laplace transform of z.
Theorem 2.6
[16] Let \(\rho >0\) and z be a piecewise continuous function on interval [0, T] and of exponential order. Then the \(\rho \)-Laplace transform of the left generalized fractional integral starting at 0 is expressed in the following form:
In the following theorem, we provided the linearity property.
Theorem 2.7
[16] If the \(\rho \)-Laplace transform of the functions \(x:[0,\infty )\rightarrow {\mathbb {R}}\) and \(y:[0,\infty )\rightarrow {\mathbb {R}}\) exists respectively for \(\tau >a_{1}\) and \(\tau >a_{2}\) and suppose \({\mathcal {L}}_{\rho }\{x(t)\}=X(\tau )\), \({\mathcal {L}}_{\rho }\{y(t)\}=Y(\tau )\). Then, the \(\rho \)-Laplace transform of \(px+qy\) exists for any constants p and q and
We need to state the \(\rho \)-convolution integral, so as to find \(\rho \)-Laplace transforms of the generalized fractional integrals and derivatives.
Definition 2.8
[16] Let x and y be piecewise continuous functions on interval [0, T] and of exponential order \(e^{a\frac{t^{\rho }}{\rho }}\). Then the \(\rho \)-convolution of x and y is defined as
In the following Lemma, we defined the commutativity of \(\rho \)-convolution of two functions.
Lemma 2.9
[16] Let x and y be piecewise continuous functions at each interval [0, T] and of exponential order. Then
and
Lemma 2.10
[16] Let \(s>0\), \(\rho >0\) and \(|\frac{\mu }{s^{\gamma }}|<1\).
-
(i)
\({\mathcal {L}}_{\rho }\{1\}(s)=\frac{1}{s}\).
-
(ii)
\({\mathcal {L}}_{\rho }\{e^{\mu \frac{t^{\rho }}{\rho }}\}(s)=\frac{1}{s-\mu }\).
-
(iii)
\({\mathcal {L}}_{\rho }\{t^{n}\}(s)=\rho ^{\frac{n}{\rho }}\frac{\Gamma (1+\frac{n}{\rho })}{s^{1+\frac{n}{\rho }}}, \; n \in {\mathbb {R}}\).
Theorem 2.11
[14] Let (W, d) be a complete metric space and S be a nonempty and closed subset of W. If a map \(\Delta :S\rightarrow S\) be a contraction, then T has specifically a unique fixed point \(w^{*} \in S\).
Theorem 2.12
[14] Let \(\mathrm {Y}\) be a Banach space and \(\Delta :\mathrm {Y}\rightarrow \mathrm {Y}\) be a continuous and compact mapping. If the set
is bounded. Then \(\Delta \) has a fixed point in \(\mathrm {Y}\).
Next, we recall the generalized Gronwall’s inequality [2].
Lemma 2.13
Let \(\gamma ,\rho >0\), \(z(t),\alpha (t)\) be nonnegative functions and \(\beta (t)\) be nondecreasing and nonnegative function, \(\beta (t)\le A\), where A is a constant. If
then we have the following inequality
In the following lemma, we present the semigroup properties.
Lemma 2.14
[17]
-
(i)
The fractional integral \({I_{t_0}^{\rho ,\gamma }}\) satisfy the semigroup property for \(\gamma ,\beta >0\)
$$\begin{aligned} ({I_{t_0}^{\rho ,\gamma }}{I_{t_0}^{\rho ,\beta }}h)(t)=({I_{t_0}^{\rho ,\gamma +\beta }}h)(t). \end{aligned}$$ -
(ii)
Let \(\gamma ,\beta >0\). The operator \({^{c}D_{t_0}^{\rho ,\gamma }}\) satisfies the semigroup property
$$\begin{aligned} ({^{c}D_{t_0}^{\rho ,\gamma }} {I_{t_0}^{\rho ,\gamma }}h)(t)=h(t), ~ ({^{c}D_{t_0}^{\rho ,\gamma }} {I_{t_0}^{\rho ,\beta }}h)(t)=({I_{t_0}^{\rho ,\beta -\gamma }}h)(t). \end{aligned}$$ -
(iii)
Let \(0<\gamma <1\), if \(h(t)\in C[t_{0},b]\) and \({I_{t_0}^{1-\gamma ,\rho }}h \in C^{1}[t_{0},b]\). Then, we have
$$\begin{aligned} ({I_{t_0}^{\rho ,\gamma }} ({^{c}D_{t_0}^{\rho ,\gamma }}h))(t)=h(t)-\frac{({I_{t_0}^{\rho ,1-\gamma }}h)(t_0)}{\Gamma (\gamma )}\left( \frac{t^{\rho }-t_{0}^{\rho }}{\rho }\right) ^{\gamma -1}, \quad \text {for all} \quad t \in (t_{0},b]. \end{aligned}$$
We present the following definitions from [21, 29].
Definition 2.15
Problem (1) is said to be UH stable if there exists a real number \(Z>0\), such that for every \(\varepsilon >0\) and for each solution \({\mathcal {W}}\in M_{n,m}\) (the class of \(n\times m\) matrices) of the inequality
there exists a solution \({\mathcal {W}}_{a}\in M_{n,m}\) of (1) satisfying
Definition 2.16
Problem (1) is said to be generalized UH–Rassias stable with respect to \(Q \in C([0,T],{\mathbb {R}})\) if there exists a constant \(u_{Q}>0\), such that for every solution \({\mathcal {W}}\in M_{n,m}\) of the inequality
there exists a solution \({\mathcal {W}}_{a}\in M_{n,m}\) of (1) satisfying
Lemma 2.17
If the matrix \((\tau ^{\gamma }I_{n}-A)\) is invertible, then the linear fractional differential equation
has solution given by
Proof
By applying the \(\rho \)-Laplace transform to both sides of (3) and using Theorem 2.5, we have
It follows that
This implies that,
\(\square \)
Lemma 2.18
Let \(\rho >0\), \(\gamma \in (0,1]\) and \(\aleph :J\times {\mathbb {R}} \times {\mathbb {R}}\rightarrow {\mathbb {R}}\), \(l:\Delta \times {\mathbb {R}} \rightarrow {\mathbb {R}}\) be continuous. Then, solution of the Cauchy problem
is given by
Proof
Let \(\alpha (t)\) satisfies (4). If \(t \in [t_0,T]\), then
Applying \( {^{c}{\mathcal {J}}_{t_0}^{\gamma ,\rho }}\) to both sides we get,
By using initial condition, we get
Next, we prove the sufficient part.
Let \(\alpha \in C_{1-\gamma ,\rho }^{\gamma }[t_0,T]\) satisfies (5), then \(^{c}D_{t_{0+}}^{\gamma ,\rho } \in C_{1-\gamma ,\rho }[t_0,T]\).
By applying the operator \(^{c}D_{t_{0+}}^{\gamma ,\rho }\) to both sides of (5), we obtain
Since,
Thus, we obtain
To show that \(\alpha (t)\) satisfies the initial condition, we apply the operator \({\mathcal {J}}_{t_0}^{1-\gamma ,\rho }\) on both sides of (5)
Now, taking the limit \(t\rightarrow t_{0}\) to both sides, we get
\(\square \)
Stability Analysis of (1)
Theorem 3.1
Let \(\gamma \in (0,1)\), \(\rho >0\), A is n-th order matrix and q(t) is specified vector. If \({\mathcal {W}}(t)\in M_{n,m}\) satisfies the following inequality
for each \(\varepsilon >0\), then there exists a solution \({\mathcal {W}}_{a}\) of problem (1) such that
Proof
Suppose
By taking the \(\rho \)-Laplace transform of (7) and using Theorem 2.5, we get
where, \({\mathcal {L}}_{\rho }\{{\mathcal {W}}_{1}(t)\}\) is the \(\rho \)-Laplace transform of the function \({\mathcal {W}}_{1}\).
Thus, from (8), we have
Set
Now, taking the \(\rho \)-Laplace transform of (10), we obtain
By Theorem 2.5 and (11), we get
Due to the fact that \({\mathcal {L}}_{\rho }\) is one-to-one infers that \({\mathcal {W}}_{a}\) is a solution of (1).
Equations (9) and (11) implies that,
This implies that,
Hence, we can say that (1) is UH stable with constant \(Z=(\frac{T^{\rho }}{\rho })^{\gamma }{\mathbb {E}}_{\gamma ,\gamma +1}(A\frac{T^{\rho }}{\rho })^{\gamma }\) for \(t\in [t_0,T]\). Note that, (1) is not UH stable, at \(t=\infty \). \(\square \)
Theorem 3.2
Suppose that \(\rho >0\), \(\gamma \in (0,1)\), \(A \in M_{n,n}\), \(P\in M_{n,m} \) and q(t) be a locally column vector containing real continuous function on \([t_0,\infty )\). If the function \({\mathcal {W}}\in {M}_{n,m}\) satisfies the following inequality
then there exists a solution \({\mathcal {W}}_{a}:[t_0,\infty )\rightarrow {\mathbb {R}}\) of problem (1) satisfying
Proof
Let
where, \({\mathcal {W}}_{1}\) is defined in (7).
Following the same procedure like in Theorem 3.1, we obtain
From (12), we conclude that (1) is generalized UH–Rassias stable on \([t_0,T]\). \(\square \)
Existence Result of Nonlinear Model (2)
We start this section by considering the following assumptions.
Suppose that
- \((\mathrm {A}_{1})\):
-
\(\aleph :J\times {\mathbb {R}}\times {\mathbb {R}}\rightarrow {\mathbb {R}}\) is continuous.
- \((\mathrm {A}_{2})\):
-
For any \(a,{\tilde{a}},b,{\tilde{b}}\in {\mathbb {R}}\), there exists a constant \({\mathcal {P}}>0\), such that
$$\begin{aligned} |\aleph (t,a,b)-\aleph (t,{\tilde{a}},{\tilde{b}})|\le {\mathcal {P}}(|a-{\tilde{a}}|+|b-{\tilde{b}}|). \end{aligned}$$ - \((\mathrm {A}_{3})\):
-
There exists a constant \(\Re >0\) such that for each \(t \in J\) and for every \(a,b \in {\mathbb {R}}\) satisfying
$$\begin{aligned} |\aleph (t,a,b)|\le \Re . \end{aligned}$$ - \((\mathrm {A}_{4})\):
-
The function \(l:\Delta \times {\mathbb {R}}\rightarrow {\mathbb {R}}\) is continuous and there exists a constant \(\mathrm {L}>0\), such that
$$\begin{aligned} |l(t,s,a)-l(t,s,{\tilde{a}})|\le \mathrm {L}_{1}|a-{\tilde{a}}|. \end{aligned}$$
We transform problem (2) into a fixed point problem.
Consider an operator \(\Psi :C(J,{\mathbb {R}})\rightarrow C(J,{\mathbb {R}})\) defined by
Our first result for the existence of solutions is based on the Schaefer’s fixed point theorem.
Theorem 4.1
Let assumptions \((\mathrm {A}_{1})\) and \((\mathrm {A}_{3})\) holds. Then, (2) has atleast one solution.
Proof
The proof is accomplish in several steps.
-
Step 1. The operator (13) is continuous.
$$\begin{aligned} |\Psi (\alpha _{n})(t)-\Psi (\alpha )(t)|&=\frac{1}{\Gamma (\gamma )}\left| \int _{t_{0}}^{t} \left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}\aleph (\tau ,\alpha _{n}(\tau ),\mathrm {L}\alpha _{n}(\tau ))\frac{d\tau }{\tau ^{1-\rho }}\right. \\&\quad -\left. \int _{t_{0}}^{t}\left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))\frac{d\tau }{\tau ^{1-\rho }}\right| \\&\le \frac{1}{\Gamma (\gamma )}\int _{t_{0}}^{t}\left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}\\&\quad \times |\aleph (\tau ,\alpha _{n}(\tau ),\mathrm {L}\alpha _{n}(\tau ))-\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))|\frac{d\tau }{\tau ^{1-\rho }}\\&\le \frac{1}{\Gamma (\gamma +1)}\left( \frac{T^{\rho }}{\rho }\right) ^{\gamma }\Vert \aleph (\cdot ,\alpha _{n},\mathrm {L}\alpha _{n})-\aleph (\cdot ,\alpha ,\mathrm {L}\alpha )\Vert _{C}. \end{aligned}$$As, \(\aleph \) is continuous function, we have
$$\begin{aligned} \Vert \aleph (\cdot ,\alpha _{n},\mathrm {L}\alpha _{n})-\aleph (\cdot ,\alpha ,\mathrm {L}\alpha )\Vert _{C}\rightarrow 0 \quad as \quad n\rightarrow \infty . \end{aligned}$$Hence, we can say that \(\Psi \) is continuous.
-
Step 2. \(\Psi \) maps bounded sets into bounded sets of \(C([t_0,T],{\mathbb {R}})\). In fact, we need to prove that for each \(\alpha \in \mathrm {M}_{\xi }=\{\alpha \in C(J,{\mathbb {R}}):\Vert \alpha \Vert \le \xi \}\) and for every \(\xi >0\) there exists a constant \(\eta >0\), such that \(\Vert \Psi (\alpha )\Vert _{c}\le \eta \). From \((\mathrm {A}_{3})\) and for any \(t \in [t_0,T]\), we have
$$\begin{aligned} |\Psi (\alpha )(t)|&\le \frac{1}{\Gamma (\gamma )}|\vartheta \rho ^{1-\gamma }(t^{\rho })^{\gamma -1}|\\&\quad +\frac{1}{\Gamma (\gamma )}\left| \int _{t_{0}}^{t}\left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1} \aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))\frac{d\tau }{\tau ^{1-\rho }}\right| \\&\le \frac{1}{\Gamma (\gamma )}|\vartheta \rho ^{1-\gamma }(t^{\rho })^{\gamma -1}|\\&\quad +\frac{1}{\Gamma (\gamma )}\int _{t_{0}}^{t}\left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1} |\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))|\frac{d\tau }{\tau ^{1-\rho }}\\&\le \frac{1}{\Gamma (\gamma )}|\vartheta \rho ^{1-\gamma }(T^{\rho })^{\gamma -1}|+\frac{\Re T^{\rho \gamma }}{\Gamma (\gamma +1)\rho ^{\gamma }}:=\eta , \end{aligned}$$which implies that
$$\begin{aligned} \Vert \Psi (\alpha )(t)\Vert _{C}\le \eta . \end{aligned}$$Hence, \(\Psi \) is bounded.
-
Step 3. The operator \(\Psi \) maps bounded sets into equicontinuous sets in \(C([t_0,T],{\mathbb {R}})\). Let \(\aleph \in \mathrm {M}_{\xi }\) with \(t_{1},t_{2} \in J\) and \(t_0\le t_{1}<t_{2}\le T\). Then
$$\begin{aligned} |\Psi (\alpha )(t_{1})-\Psi (\alpha )(t_{2})|&\le \frac{\vartheta \rho ^{1-\gamma }}{\Gamma (\gamma )}\big [(t_{1}^{\rho })^{\gamma -1}-(t_{2}^{\rho })^{\gamma -1}\big ]\\&\quad +\frac{1}{\Gamma (\gamma )}\int _{t_0}^{t_{1}}\left( \frac{t_{1}^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}|\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))|\frac{d\tau }{\tau ^{1-\rho }}\\&\quad -\frac{1}{\Gamma (\gamma )}\int _{t_0}^{t_{2}}\left( \frac{t_{2}^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1} |\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))|\frac{d\tau }{\tau ^{1-\rho }}\\&\le \frac{\vartheta \rho ^{1-\gamma }}{\Gamma (\gamma )}\big [(t_{1}^{\rho })^{\gamma -1}-(t_{2}^{\rho })^{\gamma -1}\big ]\\&\quad +\frac{\Re }{\Gamma (\gamma )}\left( \int _{t_0}^{t_{1}}\left[ \left( \frac{t_{1}^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1} -\left( \frac{t_{2}^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}\right] \frac{d\tau }{\tau ^{1-\rho }}\right. \\&\quad +\left. \int _{t_{1}}^{t_{2}}\left( \frac{t_{2}^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1}\frac{d\tau }{\tau ^{1-\rho }}\right) \\&\le \frac{\vartheta \rho ^{1-\gamma }}{\Gamma (\gamma )}\big [(t_{1}^{\rho })^{\gamma -1}-(t_{2}^{\rho })^{\gamma -1}\big ]\\&\quad +\frac{\Re }{\Gamma (\gamma +1)\rho ^{\gamma }}\big (t_{2}^{\gamma \rho }-t_{1}^{\gamma \rho }-2(t_{2}^{\rho }-t_{1}^{\rho })^{\gamma }\big ). \end{aligned}$$As, \(t_{1}\rightarrow t_{2}\), then the right hand side of the above inequality tends to zero. Consequently, \(\Psi \) is equicontinuous. As a consequence of Steps 1–3, together with the Arzela–Ascoli theorem, we can conclude that \(\Psi \) is completely continuous.
-
Step 4. Next, we show that the set \(\Lambda =\{\alpha \in C(J,{\mathbb {R}}):\alpha =\kappa \Omega (\alpha )\;\text {for some}\; 0<\kappa <1\}\) is bounded. Let \(\alpha \in \Lambda \) and for each \(t \in J\), we have:
$$\begin{aligned} |\alpha (t)|&\le \kappa \left[ \frac{\vartheta }{\Gamma (\gamma )}\left( \frac{t^{\rho }}{\rho }\right) ^{\gamma -1} +\frac{1}{\Gamma (\gamma )}\int _{t_0}^{t}\left( \frac{t^{\rho }-\tau ^{\rho }}{\rho }\right) ^{\gamma -1} |\aleph (\tau ,\alpha (\tau ),\mathrm {L}\alpha (\tau ))|\frac{d\tau }{\tau ^{1-\rho }}\right] \\&\le \frac{\vartheta }{\Gamma (\gamma )}\left( \frac{t^{\rho }}{\rho }\right) ^{\gamma -1}+\frac{\Re T^{\rho \gamma }}{\Gamma (\gamma +1)\rho ^{\gamma }}:={\mathcal {Q}}, \end{aligned}$$
which implies that,
Thus, the set \(\Lambda \) is bounded. \(\square \)
Our next analysis is to show that the operator \(\Psi \) has a unique fixed point.
Theorem 4.2
Assume that \((A_{1}),(A_{2})\) and \((A_{4})\) holds. Then, the problem (2) has a unique solution on \([t_0,T]\), provided that
Proof
By the use of Banach fixed point theorem, we can prove that the operator defined by (13) has a fixed point.
For \(\alpha _{1},\alpha _{2} \in C(J,{\mathbb {R}})\) and \(t \in [t_0,T]\), we have
On further simplifications, we get
Hence, \(\Psi \) is a contraction. As an application of Banach fixed point theorem, we conclude that \(\Psi \) has a fixed point which is the solution of (2). \(\square \)
Stability Analysis for Nonlinear Model
In this section, we are interested in the stability concepts from the Ulam’s point of view for the problem (2).
We consider the following definitions.
Definition 5.1
Problem (2) is said to be UH stable, if there exists a constant \(M>0\) such that for every \(\xi >0\) and for each solution \(\alpha \in C(J,{\mathbb {R}})\) of the inequality
there exists a solution \(\alpha ^{*}\in C(J,{\mathbb {R}})\) of the problem (2) satisfying
Definition 5.2
Problem (2) is said to be generalized UH stable, if there is \(\phi \in C({\mathbb {R}}_{+},{\mathbb {R}}_{+})\), \(\phi (0)=0\) such that for every \(\xi >0\) and for each solution \(\alpha \in C(J,{\mathbb {R}})\) satisfying
there exists a solution \(\alpha ^{*}\in C(J,{\mathbb {R}})\) of the problem (2) with
Definition 5.3
Problem (2) is said to be UH–Rassias stable with respect to \(\Theta \in C(J,{\mathbb {R}})\), if there exists a constant \(M_{\Theta }>0\) such that for every \(\xi >0\) and for each solution \(\alpha \in C(J,{\mathbb {R}})\) of the inequality
there exists a solution \(\alpha ^{*}\in C(J,{\mathbb {R}})\) of the problem (2) satisfying
Definition 5.4
Problem (2) is said to be generalized UH–Rassias stable with respect to \(\Theta \in C(J,{\mathbb {R}})\), if there exists a constant \(M_{\Theta }>0\) and for each solution \(\alpha \in C(J,{\mathbb {R}})\) of the inequality
there exists a solution \(\alpha ^{*}\in C(J,{\mathbb {R}})\) of the problem (2) satisfying
Remark 5.5
It is clear that
Remark 5.6
We say that \(\alpha \in C(J,{\mathbb {R}})\) is a solution of the inequality (15), if there exists a function \(\Theta \in C(J,{\mathbb {R}})\) which depends on \(\alpha \) such that
-
(I)
\(|\Theta (t)|\le \delta , \; t \in J;\)
-
(II)
\(^{c}D_{t_{0+}}^{\gamma ,\rho }\alpha (t)=\aleph (t,\alpha (t),\mathrm {L}\alpha (t))+\Theta (t), \; t\in J\).
Similar remarks can be stated for the inequalities (17) and (18).
Remark 5.7
Let \(\rho >0\), \(\alpha \in (0,1]\), if \(\alpha \in C(J,{\mathbb {R}})\) is a solution of the inequality (15), then we have
Proof
In fact, by (II) of Remark 5.6, we have
Then,
It follows that
One can state similar remarks for the solutions of the inequalities (17) and (18). \(\square \)
Before stating our main result, we need the following assumption.
- \((\mathrm {A}_{5})\):
-
Suppose that \(\Theta \) be nondecreasing function and there exists a constant \(\zeta _{\Theta }>0\) such that
$$\begin{aligned} \int _{t_0}^{t}(t^{\rho }-s^{\rho })^{\gamma -1}(\tau ^{\rho })' \Theta (\tau )d\tau \le \zeta _{\Theta } \Theta , \end{aligned}$$(19)for all \(t \in [t_0,T]\).
Theorem 5.8
Let the assumptions \((A_{1}),(A_{2}),(A_{4})\), and \((A_5)\) hold. Then, (2) is said to be generalized Ulam–Hyers–Rassias stable on J.
Proof
Suppose that \(\alpha \in C(J,{\mathbb {R}})\) be the solution of the inequality (18). From Theorem 4.2, \(\alpha ^{*} \in C(J,{\mathbb {R}})\) be the unique solution of the following
We write the solution of (20) as
Now, integrating the inequality (18) and by condition \((A_{5})\), we have
Accordingly,
Hence, by utilizing Lemma 2.13, we acquire
Here we set,
Hence, the proposed problem (2) is generalized UH–Rassias stable with respect to \(\Theta (t)\) on \([t_0,T]\). \(\square \)
Example
This section is devoted to present the application of our main results.
Example 6.1
Consider the following nonlinear fractional integro-differential equation with Liouville–Caputo derivative
Note that, \(\gamma =\frac{1}{3}\), \(\rho =3\), \(\aleph (t,\alpha (t),\int _{0}^{t}l(s,\alpha (s))ds)=\frac{1}{2\sqrt{4+t^{2}}}\bigg [\frac{|\alpha (t)|}{1+|\alpha (t)|}+\int _{0}^{t}e^{3s}\frac{|\alpha (s)|}{1+|\alpha (s)|}ds\bigg ]\) and \((t,\alpha ) \in [0,2] \times {\mathbb {R}}\).
Here, \({\mathcal {P}}=\frac{1}{4}\) and \(\mathrm {L}_{1}=\frac{1}{3}\). Then \(\frac{{\mathcal {P}}(1+\mathrm {L}_{1})T^{\rho \gamma }}{\Gamma (\gamma +1)\rho ^{\rho }}=0.5175<1\).
By Theorem 4.2, we can say that problem (21) has a unique solution. Next, we will check the assumption \((A_{5})\).
Let \(\Theta (t)=e^{t}\), \(\zeta _{\Theta }=\frac{2}{3}>0\) for \(t \in [0,2]\).
As a consequence, we can say that problem (21) is generalized UH–Rassias stable.
Conclusion
The \(\rho \)-Laplace transform is sufficient to deal with linear problems. In fact, we utilize the modified Laplace transform method to investigate the UH and generalized UH–Rassias stability of Liouville–Caputo fractional differential equations. The best tools to work with nonlinear models are the fixed point approaches and Gronwall’s inequality. Hence, in view of Schaefer’s and Banach fixed point theorem, we procured our qualitative analysis i.e., existence and uniqueness of solutions of problem (2). To achieve the stability results for nonlinear fractional integro-differential equations, we provided some hypothesis along with the extended Gronwall’s inequality. To justify our work, an application is provided.
Availability of data and materials
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.
References
Abdeljawad, T., Alzabut, J.: The q-fractional analogue for Gronwall-type inequality. J. Funct. Space 2013, 1–7 (2013)
Adjabi, Y., Jarad, F., Abdeljawad, T.: On generalized fractional operators and a Gronwall type inequality with applications. Filomat 31, 5457–5473 (2017)
Ahmad, M., Zada, A., Alzabut, J.: Hyers–Ulam stability of a coupled system of fractional differential equations of Hilfer–Hadamard type. Demonstr. Math. 52, 283–295 (2019)
Ahmad, M., Jiang, J., Zada, A., Ali, Z., Fu, Z., Xu, J.: Hyers–Ulam–Mittag–Leffler stability for a system of fractional neutral differential equations. Disrete Dyn. Nat. Soc. 5, 1–10 (2020)
Ahmad, M., Jiang, J., Zada, A., Shah, S.O., Xu, J.: Analysis of coupled system of implicit fractional differential equations involving Katugampola–Caputo fractional derivative. Complexity 4, 1–11 (2020)
Ahmad, M., Zada, A., Dong, W., Xu, J.: Stability analysis of a nonlocal fractional impulsive coupled evolution differential equation. J. Appl. Anal. Comput. 11(1), 138–160 (2021)
Altun, Y.: Further results on the asymptotic stability of Riemann–Liouville fractional neutral systems with variable delays. Adv. Differ. Equ. 437, 1–13 (2019)
Atangana, A., Baleanu, D.: New fractional derivatives with nonlocal and non-singular kernel: theory and application to heat transfer model. Therm. Sci. 20, 763–769 (2016)
Baitiche, Z., Derbazi, C., Matar, M.M.: Ulam stability for nonlinear-Langevin fractional differential equations involving two fractional orders in the \(\psi \)-Caputo sense. Appl. Anal. 1–17 (2021)
Corduneanu, C.: Principles of Differential and Integral Equations. Allyn and Bacon, Boston (1971)
Derbazi, C., Baitiche, Z.: Uniqueness and Ulam–Hyers–Mittag–Leffler stability results for the delayed fractional multi-terms differential equation involving the \(\phi \)-Caputo fractional derivative. arXiv:2012.10233v1 [math.FA] (2020)
Derbazi, C., Baitiche, Z., Benchohra, M., Cabada, A.: Initial value problem for nonlinear fractional differential equations with \(\psi \)-Caputo derivative via monotone iterative technique. Axioms 1–13 (2020)
Derbazi, C., Baitiche, Z., Benchohra, M., N’guerekata, G.: Existence, uniqueness, approximation of solutions and \(E_{\alpha }\)-Ulam stability results for a class of nonlinear fractional differential equations involving \(\psi \)-Caputo derivative with initial conditions. Mathematica Moravica 25(1), 1–30 (2021)
Granas, A., Dugundji, J.: Fixed Point Theory. Springer, New York (2003)
Hyers, D.H.: On the stability of the linear functional equation. Proc. Natl. Acad. Sci. USA 27, 222–224 (1941)
Jarad, F., Abdeljawad, T.: A modified Laplace transform for certain generalized fractional operators. Results Nonlinear Anal. 1, 88–98 (2018)
Jarad, F., Abdeljawad, T., Baleanu, D.: On the generalized fractional derivatives and their Caputo modification. J. Nonlinear Sci. Appl. 10, 2607–2619 (2017)
Katugampola, U.N.: New approach to generalized fractional integral. Appl. Math. Comput. 218, 860–865 (2011)
Katugampola, U.N.: A new approach to generalized fractional derivatives. Bull. Math. Anal. Appl. 6, 1–15 (2014)
Li, Y., Chen, Y., Podlubny, I.: Stability of fractional-order nonlinear dynamic systems: Lyapunov direct method and generalized Mittag–Leffler stability. Comput. Math. Appl. 5, 1810–1821 (2010)
Liu, K., Fečkan, M., O’Regan, D., Wang, J.: Hyers–Ulam stability and existence of solutions for differential equations with Caputo–Fabrizio fractional derivative. Open Math. 333, 1–14 (2019)
Liu, K., Fečkan, M., Wang, J.: Hyers–Ulam stability and existence of solutions to the generalized Liouville–Caputo fractional differential equations. Symmetry 12, 1–18 (2020)
Liu, K., Wang, J., Zhou, Y., O’Regan, D.: Hyers–Ulam stability and existence of solutions for fractional differential equations with Mittag–Leffler kernel. Chaos Solitons Fractals 132, 1–8 (2020)
Luo, D., Zada, A., Shaleena, S., Ahmad, M.: Analysis of a coupled system of fractional differential equations with non-separated boundary conditions. Adv. Differ. Equ. 590, 1–24 (2020)
Rezaei, H., Jung, S.M., Rassias, Th.M.: Laplace transform and Hyers–Ulam stability of linear differential equations. J. Math. Anal. Appl. 403, 244–251 (2013)
Sene, N., Srivastava, G.: Generalized Mittag–Leffler input stability of the fractional differential equations. Symmetry 608, 1–12 (2019)
Ulam, S.M.: A Collection of the Mathematical Problems. Interscience Publishers, New York (1960)
Wang, C., Xu, T.: Hyers–Ulam stability of fractional linear differential equations involving Caputo fractional derivatives. Appl. Math. 60, 383–393 (2015)
Wang, J., Lv, L., Zhou, Y.: Ulam stability and data dependence for fractional differential equations with Caputo derivative. Electron. J. Qual. Theory Differ. Equ. 63, 1–10 (2011)
Ye, H., Gao, J., Ding, Y.: A generalized Gronwall inequality and its application to a fractional differential equation. J. Math. Anal. Appl. 328, 1075–1081 (2007)
Zada, A., Mashal, A.: Stability analysis of nth order nonlinear impulsive differential equations in quasi-Banach space. Numer. Funct. Anal. Optim. 41, 294–321 (2020)
Zada, A., Shaleena, S., Li, T.: Stability analysis of higher order nonlinear differential equations in \(\beta \) normed spaces. Math. Methods Appl. Sci. 42, 1151–1166 (2019)
Zada, A., Alzabut, J., Waheed, H., Popa, I.L.: Ulam–Hyers stability of impulsive integrodifferential equations with Riemann-Liouville boundary conditions. Adv. Differ. Equ. 64, 1–50 (2020)
Funding
Not applicable.
Author information
Authors and Affiliations
Contributions
All the authors contributed equally and considerably in this paper.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Zada, A., Shaleena, S. & Ahmad, M. Analysis of Solutions of the Integro-Differential Equations with Generalized Liouville–Caputo Fractional Derivative by \(\rho \)-Laplace Transform. Int. J. Appl. Comput. Math 8, 116 (2022). https://doi.org/10.1007/s40819-022-01275-8
Accepted:
Published:
DOI: https://doi.org/10.1007/s40819-022-01275-8
Keywords
- Generalized Liouville–Caputo fractional differential equations
- \(\rho \)-Laplace transform
- Mittag–Leffler function
- Fixed point approach
- Hyers–Ulam stability