The aim of this paper is to examine exchange rate volatility using GARCH models with a new innovation distribution, the Normal Tempered Stable. We estimated daily exchange rate volatility using different distributions (Normal, Student, NIG) in order to specify the performed model. In addition, a forecasting analysis is performed to check which distribution reveals the best out-of-sample results. We found that the estimated parameters of GARCH-NTS model outperform the GARCH-N and GARCH-t ones for all currencies. Besides, we asserted that GARCH-NTS and EGARCH-NTS are the preferred models in terms of out-of sample forecasting accuracy. Our results indicating the performance of GARCH models with NTS distribution contribute to increase the accuracy of risk measures which is very important for international traders and investors.
Citation: Sahar Charfi, Farouk Mselmi. Modeling exchange rate volatility: application of GARCH models with a Normal Tempered Stable distribution[J]. Quantitative Finance and Economics, 2022, 6(2): 206-222. doi: 10.3934/QFE.2022009
[1] | Ishfaq Mallah, Idris Ahmed, Ali Akgul, Fahd Jarad, Subhash Alha . On ψ-Hilfer generalized proportional fractional operators. AIMS Mathematics, 2022, 7(1): 82-103. doi: 10.3934/math.2022005 |
[2] | Mohammed A. Almalahi, Satish K. Panchal, Fahd Jarad, Mohammed S. Abdo, Kamal Shah, Thabet Abdeljawad . Qualitative analysis of a fuzzy Volterra-Fredholm integrodifferential equation with an Atangana-Baleanu fractional derivative. AIMS Mathematics, 2022, 7(9): 15994-16016. doi: 10.3934/math.2022876 |
[3] | Dumitru Baleanu, Babak Shiri . Generalized fractional differential equations for past dynamic. AIMS Mathematics, 2022, 7(8): 14394-14418. doi: 10.3934/math.2022793 |
[4] | Marimuthu Mohan Raja, Velusamy Vijayakumar, Anurag Shukla, Kottakkaran Sooppy Nisar, Wedad Albalawi, Abdel-Haleem Abdel-Aty . A new discussion concerning to exact controllability for fractional mixed Volterra-Fredholm integrodifferential equations of order r∈(1,2) with impulses. AIMS Mathematics, 2023, 8(5): 10802-10821. doi: 10.3934/math.2023548 |
[5] | Sumbal Ahsan, Rashid Nawaz, Muhammad Akbar, Saleem Abdullah, Kottakkaran Sooppy Nisar, Velusamy Vijayakumar . Numerical solution of system of fuzzy fractional order Volterra integro-differential equation using optimal homotopy asymptotic method. AIMS Mathematics, 2022, 7(7): 13169-13191. doi: 10.3934/math.2022726 |
[6] | Shuang-Shuang Zhou, Saima Rashid, Saima Parveen, Ahmet Ocak Akdemir, Zakia Hammouch . New computations for extended weighted functionals within the Hilfer generalized proportional fractional integral operators. AIMS Mathematics, 2021, 6(5): 4507-4525. doi: 10.3934/math.2021267 |
[7] | Muhammad Tariq, Sotiris K. Ntouyas, Hijaz Ahmad, Asif Ali Shaikh, Bandar Almohsen, Evren Hincal . A comprehensive review of Grüss-type fractional integral inequality. AIMS Mathematics, 2024, 9(1): 2244-2281. doi: 10.3934/math.2024112 |
[8] | Sajid Iqbal, Muhammad Samraiz, Gauhar Rahman, Kottakkaran Sooppy Nisar, Thabet Abdeljawad . Some new Grüss inequalities associated with generalized fractional derivative. AIMS Mathematics, 2023, 8(1): 213-227. doi: 10.3934/math.2023010 |
[9] | Lakhlifa Sadek, Tania A Lazǎr . On Hilfer cotangent fractional derivative and a particular class of fractional problems. AIMS Mathematics, 2023, 8(12): 28334-28352. doi: 10.3934/math.20231450 |
[10] | Saima Rashid, Abdulaziz Garba Ahmad, Fahd Jarad, Ateq Alsaadi . Nonlinear fractional differential equations and their existence via fixed point theory concerning to Hilfer generalized proportional fractional derivative. AIMS Mathematics, 2023, 8(1): 382-403. doi: 10.3934/math.2023018 |
The aim of this paper is to examine exchange rate volatility using GARCH models with a new innovation distribution, the Normal Tempered Stable. We estimated daily exchange rate volatility using different distributions (Normal, Student, NIG) in order to specify the performed model. In addition, a forecasting analysis is performed to check which distribution reveals the best out-of-sample results. We found that the estimated parameters of GARCH-NTS model outperform the GARCH-N and GARCH-t ones for all currencies. Besides, we asserted that GARCH-NTS and EGARCH-NTS are the preferred models in terms of out-of sample forecasting accuracy. Our results indicating the performance of GARCH models with NTS distribution contribute to increase the accuracy of risk measures which is very important for international traders and investors.
Recently, fractional calculus has attained assimilated bounteous flow and significant importance due to its rife utility in the areas of technology and applied analysis. Fractional derivative operators have given a new rise to mathematical models such as thermodynamics, fluid flow, mathematical biology, and virology, see [1,2,3]. Previously, several researchers have explored different concepts related to fractional derivatives, such as Riemann-Liouville, Caputo, Riesz, Antagana-Baleanu, Caputo-Fabrizio, etc. As a result, this investigation has been directed at various assemblies of arbitrary order differential equations framed by numerous analysts, (see [4,5,6,7,8,9,10]). It has been perceived that the supreme proficient technique for deliberating such an assortment of diverse operators that attracted incredible presentation in research-oriented fields, for example, quantum mechanics, chaos, thermal conductivity, and image processing, is to manage widespread configurations of fractional operators that include many other operators, see the monograph and research papers [11,12,13,14,15,16,17,18,19,20,21,22].
In [23], the author proposed a novel idea of fractional operators, which is called GPF operator, that recaptures the Riemann-Liouville fractional operators into a solitary structure. In [24], the authors analyzed the existence of the FDEs as well as demonstrated the uniqueness of the GPF derivative by utilizing Kransnoselskii's fixed point hypothesis and also dealt with the equivalency of the mixed type Volterra integral equation.
Fractional calculus can be applied to a wide range of engineering and applied science problems. Physical models of true marvels frequently have some vulnerabilities which can be reflected as originating from various sources. Additionally, fuzzy sets, fuzzy real-valued functions, and fuzzy differential equations seem like a suitable mechanism to display the vulnerabilities marked out by elusiveness and dubiousness in numerous scientific or computer graphics of some deterministic certifiable marvels. Here we broaden it to several research areas where the vulnerability lies in information, for example, ecological, clinical, practical, social, and physical sciences [25,26,27].
In 1965, Zadeh [28] proposed fuzziness in set theory to examine these issues. The fuzzy structure has been used in different pure and applied mathematical analyses, such as fixed-point theory, control theory, topology, and is also helpful for fuzzy automata and so forth. In [29], authors also broadened the idea of a fuzzy set and presented fuzzy functions. This concept has been additionally evolved and the bulk of the utilization of this hypothesis has been deliberated in [30,31,32,33,34,35] and the references therein. The concept of HD has been correlated with fuzzy Riemann-Liouville differentiability by employing the Hausdorff measure of non-compactness in [36,37].
Numerous researchers paid attention to illustrating the actual verification of certain fuzzy integral equations by employing the appropriate compactness type assumptions. Different methodologies and strategies, in light of HD or generalized HD (see [38]) have been deliberated in several credentials in the literature (see for instance [39,40,41,42,43,44,45,46,47,48,49]) and we presently sum up quickly a portion of these outcomes. In [50], the authors proved the existence of solutions to fuzzy FDEs considering Hukuhara fractional Riemann-Liouville differentiability as well as the uniqueness of the aforesaid problem. In [51,52], the authors investigated the generalized Hukuhara fractional Riemann-Liouville and Caputo differentiability of fuzzy-valued functions. Bede and Stefanini [39] investigated and discovered novel ideas for fuzzy-valued mappings that correlate with generalized differentiability. In [43], Hoa introduced the subsequent fuzzy FDE with order ϑ∈(0,1):
{(cDϑσ+1Φ)(ζ)=F(ζ,Φ(ζ)),Φ(σ1)=Φ0∈E, | (1.1) |
where a fuzzy function is F:[σ1,σ2]×E→E with a nontrivial fuzzy constant Φ0∈E. The article addressed certain consequences on clarification of the fractional fuzzy differential equations and showed that the aforesaid equations in both cases (differential/integral) are not comparable in general. A suitable assumption was provided so that this correspondence would be effective. Hoa et al. [53] proposed the Caputo-Katugampola FDEs fuzzy set having the initial condition:
{(cDϑ,ρσ+1Φ)(ζ)=F(ζ,Φ(ζ)),Φ(σ1)=Φ0, | (1.2) |
where 0<σ1<ζ≤σ2, cDϑ,ρσ+1 denotes the fuzzy Caputo-Katugampola fractional generalized Hukuhara derivative and a fuzzy function is F:[σ1,σ2]×E→E. An approach of continual estimates depending on generalized Lipschitz conditions was employed to discuss the actual as well as the uniqueness of the solution. Owing to the aforementioned phenomena, in this article, we consider a novel fractional derivative (merely identified as Hilfer GPF-derivative). Consequently, in the framework of the proposed derivative, we establish the basic mathematical tools for the investigation of GPF-FFHD which associates with a fractional order fuzzy derivative. We investigated the actuality and uniqueness consequences of the clarification to a fuzzy fractional IVP by employing GPF generalized HD by considering an approach of continual estimates via generalized Lipschitz condition. Moreover, we derived the FVFIE using a generalized fuzzy GPF derivative is presented. Finally, we demonstrate the problems of actual and uniqueness of the clarification of this group of equations. The Hilfer-GPF differential equation is presented as follows:
{Dϑ,q,βσ+1Φ(ζ)=F(ζ,Φ(ζ)),ζ∈[σ1,T],0≤σ1<TI1−γ,βσ1Φ(σ1)=m∑j=1RjΦ(νj),ϑ≤γ=ϑ+q−ϑq,νj∈(σ1,T], | (1.3) |
where Dϑ,q,βσ+1(.) is the Hilfer GPF-derivative of order ϑ∈(0,1),I1−γ,βσ1(.) is the GPF integral of order 1−γ>0,Rj∈R, and a continuous function F:[σ1,T]×R→R with νj∈[σ1,T] fulfilling σ<ν1<...<νm<T for j=1,...,m. To the furthest extent that we might actually know, nobody has examined the existence and uniqueness of solution (1.3) regarding FVFIEs under generalized fuzzy Hilfer-GPF-HD with fuzzy initial conditions. An illustrative example of fractional-order in the complex domain is proposed and provides the exact solution in terms of the Fox-Wright function.
The following is the paper's summary. Notations, hypotheses, auxiliary functions, and lemmas are presented in Section 2. In Section 3, we establish the main findings of our research concerning the existence and uniqueness of solutions to Problem 1.3 by means of the successive approximation approach. We developed the fuzzy GPF Volterra-Fredholm integrodifferential equation in Section 4. Section 5 consists of concluding remarks.
Throughout this investigation, E represents the space of all fuzzy numbers on R. Assume the space of all Lebsegue measureable functions with complex values F on a finite interval [σ1,σ2] is identified by χrc(σ1,σ2) such that
‖F‖χrc<∞,c∈R,1≤r≤∞. |
Then, the norm
‖F‖χrc=(σ2∫σ1|ζcF(ζ)|rdζζ)1/r∞. |
Definition 2.1. ([53]) A fuzzy number is a fuzzy set Φ:R→[0,1] which fulfills the subsequent assumptions:
(1) Φ is normal, i.e., there exists ζ0∈R such that Φ(ζ0)=1;
(2) Φ is fuzzy convex in R, i.e, for δ∈[0,1],
Φ(δζ1+(1−δ)ζ2)≥min{Φ(ζ1),Φ(ζ2)}foranyζ1,ζ2∈R; |
(3) \Phi is upper semicontinuous on \mathbb{R};
(4) [z]^{0} = cl\big\{z_{1}\in\mathbb{R}\, \vert\, \Phi(z_{1}) > 0\big\} is compact.
\mathcal{C}\big([\sigma_{1}, \sigma_{2}], \mathfrak{E}\big) indicates the set of all continuous functions and set of all absolutely continuous fuzzy functions signifys by \mathcal{AC}\big([\sigma_{1}, \sigma_{2}], \mathfrak{E}\big) on the interval [\sigma_{1}, \sigma_{2}] having values in \mathfrak{E}.
Let \gamma\in(0, 1), we represent the space of continuous mappings by
\begin{eqnarray*} \mathcal{C}_{\gamma}[\sigma_{1},\sigma_{2}] = \big\{\mathcal{F}:(\sigma_{1},\sigma_{2}]\rightarrow \mathfrak{E}:e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{1-\gamma}\mathcal{F}(\zeta)\in\mathcal{C}[\sigma_{1},\sigma_{2}]\big\}. \end{eqnarray*} |
Assume that a fuzzy set \Phi:\mathbb{R}\mapsto[0, 1] and all fuzzy mappings \Phi:[\sigma_{1}, \sigma_{2}]\rightarrow \mathfrak{E} defined on L\big([\sigma_{1}, \sigma_{2}], \mathfrak{E}\big) such that the mappings \zeta\rightarrow \bar{\mathcal{D}}_{0}[\Phi(\zeta), \hat{0}] lies in L_{1}[\sigma_{1}, \sigma_{2}].
There is a fuzzy number \Phi on \mathbb{R}, we write [\Phi]^{\check{q}} = \big\{z_{1}\in\mathbb{R}\, \vert\, \Phi(z_{1})\geq \check{q}\big\} the \check{q} -level of \Phi, having \check{q}\in(0, 1].
From assertions (1) to (4); it is observed that the \check{q} -level set of \Phi\in\mathfrak{E}, [\Phi]^{\check{q}} is a nonempty compact interval for any \check{q}\in(0, 1]. The \check{q} -level of a fuzzy number \Phi is denoted by \big[\underline{\Phi}(\check{q}), \bar{\Phi}(\check{q})\big].
For any \delta\in\mathbb{R} and \Phi_{1}, \Phi_{2}\in\mathfrak{E}, then the sum \Phi_{1}+\Phi_{2} and the product \delta\Phi_{1} are demarcated as: [\Phi_{1}+\Phi_{2}]^{\check{q}} = [\Phi_{1}]^{\check{q}}+[\Phi_{2}]^{\check{q}} and [\delta.\Phi_{1}]^{\check{q}} = \delta[\Phi_{1}]^{\check{q}}, for all \check{q}\in[0, 1], where [\Phi_{1}]^{\check{q}}+[\Phi_{2}]^{\check{q}} is the usual sum of two intervals of \mathbb{R} and \delta[\Phi_{1}]^{\check{q}} is the scalar multiplication between \delta and the real interval.
For any \Phi\in\mathfrak{E}, the diameter of the \check{q} -level set of \Phi is stated as diam[\mu]^{\check{q}} = \bar{\mu}(\check{q})-\underline{\mu}(\check{q}).
Now we demonstrate the notion of Hukuhara difference of two fuzzy numbers which is mainly due to [54].
Definition 2.2. ([54]) Suppose \Phi_{1}, \Phi_{2}\in\mathfrak{E}. If there exists \Phi_{3}\in\mathfrak{E} such that \Phi_{1} = \Phi_{2}+\Phi_{3}, then \Phi_{3} is known to be the Hukuhara difference of \Phi_{1} and \Phi_{2} and it is indicated by \Phi_{1}\ominus\Phi_{2}. Observe that \Phi_{1}\ominus\Phi_{2}\neq\Phi_{1}+(-)\Phi_{2}.
Definition 2.3. ([54]) We say that {\bar{\mathcal{D}_{0}}}[\Phi_{1}, \Phi_{2}] is the distance between two fuzzy numbers if
\begin{eqnarray*} {\bar{\mathcal{D}_{0}}}[\Phi_{1},\Phi_{2}] = \sup\limits_{\check{q}\in[0,1]}\mathcal{H}\Big([\Phi_{1}]^{\check{q}},[\Phi_{2}]^{\check{q}}\Big),\quad \forall \Phi_{1},\Phi_{2}\in\mathfrak{E}, \end{eqnarray*} |
where the Hausdroff distance between [\Phi_{1}]^{\check{q}} and [\Phi_{2}]^{\check{q}} is defined as
\mathcal{H}\Big([\Phi_{1}]^{\check{q}},[\Phi_{2}]^{\check{q}}\Big) = \max\big\{\vert\underline{\Phi}(\check{q})-\bar{\Phi}(\check{q})\vert,\vert\bar{\Phi}(\check{q})-\underline{\Phi}(\check{q})\vert\big\}. |
Fuzzy sets in \mathfrak{E} is also refereed as triangular fuzzy numbers that are identified by an ordered triple \Phi = (\sigma_{1}, \sigma_{2}, \sigma_{3})\in\mathbb{R}^{3} with \sigma_{1}\leq \sigma_{2}\leq \sigma_{3} such that [\Phi]^{\check{q}} = [\underline{\Phi}(\check{q}), \bar{\Phi}(\check{q})] are the endpoints of \check{q} -level sets for all \check{q}\in[0, 1], where \underline{\Phi}(\check{q}) = \sigma_{1}+(\sigma_{2}-\sigma_{1})\check{q} and \bar{\Phi}(\check{q}) = \sigma_{3}-(\sigma_{3}-\sigma_{2})\check{q}.
Generally, the parametric form of a fuzzy number \Phi is a pair [\Phi]^{\check{q}} = [\underline{\Phi}(\check{q}), \bar{\Phi}(\check{q})] of functions \underline{\Phi}(\check{q}), \bar{\Phi}(\check{q}), \check{q}\in[0, 1], which hold the following assumptions:
(1) \underline{\mu}(\check{q}) is a monotonically increasing left-continuous function;
(2) \bar{\mu}(\check{q}) is a monotonically decreasing left-continuous function;
(3) \underline{\mu}(\check{q})\leq\bar{\mu}(\check{q}), \, \check{q}\in[0, 1].
Now we mention the generalized Hukuhara difference of two fuzzy numbers which is proposed by [38].
Definition 2.4. ([38]) The generalized Hukuhara difference of two fuzzy numbers \Phi_{1}, \Phi_{2}\in\mathfrak{E} ( \mathfrak{g}H -difference in short) is stated as follows
\begin{eqnarray*} \Phi_{1}\ominus_{\mathfrak{g}H}\Phi_{2} = \Phi_{3} \quad\Leftrightarrow\quad \Phi_{1} = \Phi_{2}+\Phi_{3}\quad or\quad\Phi_{2} = \Phi_{1}+(-1)\Phi_{3}. \end{eqnarray*} |
A function \Phi:[\sigma_{1}, \sigma_{2}]\rightarrow \mathfrak{E} is said to be \mathfrak{d} -increasing ( \mathfrak{d} -decreasing) on [\sigma_{1}, \sigma_{2}] if for every \check{q}\in[0, 1]. The function \zeta\rightarrow \, \, diam[\Phi(\zeta)]^{\check{q}} is nondecreasing (nonincreasing) on [\sigma_{1}, \sigma_{2}] . If \Phi is \mathfrak{d} -increasing or \mathfrak{d} -decreasing on [\sigma_{1}, \sigma_{2}] , then we say that \Phi is \mathfrak{d} -monotone on [\sigma_{1}, \sigma_{2}].
Definition 2.5. ([39])The generalized Hukuhara derivative of a fuzzy-valued function \mathcal{F}:(\sigma_{1}, \sigma_{2})\rightarrow \mathfrak{E} at \zeta_{0} is defined as
\begin{eqnarray*} \mathcal{F}_{\mathfrak{g}H}^{\prime}(\zeta_{0}) = \lim\limits_{h\rightarrow 0}\frac{\mathcal{F}(\zeta_{0}+h)\ominus_{\mathfrak{g}H}\mathcal{F}(\zeta_{0})}{h}, \end{eqnarray*} |
if (\mathcal{F})_{\mathfrak{g}H}^{\prime}(\zeta_{0})\in\mathfrak{E}, we say that \mathcal{F} is generalized Hukuhara differentiable ( \mathfrak{g}H -differentiable) at \zeta_{0}.
Moreover, we say that \mathcal{F} is [(i)-\mathfrak{g}H] -differentiable at \zeta_{0} if
\begin{eqnarray} [\mathcal{F}_{\mathfrak{g}H}^{\prime}(\zeta_{0})]^{\check{q}}&& = \Bigg[\bigg[\lim\limits_{h\rightarrow 0}\frac{\underline{\mathcal{F}}(\zeta_{0}+h)\ominus_{\mathfrak{g}H}\underline{\mathcal{F}}(\zeta_{0})}{h}\bigg]^{\check{q}} ,\bigg[\lim\limits_{h\rightarrow 0}\frac{\bar{\mathcal{F}}(\zeta_{0}+h)\ominus_{\mathfrak{g}H}\bar{\mathcal{F}}(\zeta_{0})}{h}\bigg]^{\check{q}} \Bigg] \\&& = \big[(\underline{\mathcal{F}})^{\prime}(\check{q},\zeta_{0}),(\bar{\mathcal{F}})^{\prime}(\check{q},\zeta_{0})\big], \end{eqnarray} | (2.1) |
and that \mathcal{F} is [(ii)-\mathfrak{g}H] -differentiable at \zeta_{0} if
\begin{eqnarray} [\mathcal{F}_{\mathfrak{g}H}^{\prime}(\zeta_{0})]^{\check{q}} = \big[(\bar{\mathcal{F}})^{\prime}(\check{q},\zeta_{0}),(\underline{\mathcal{F}})^{\prime}(\check{q},\zeta_{0})\big]. \end{eqnarray} | (2.2) |
Definition 2.6. ([49]) We state that a point \zeta_{0}\in(\sigma_{1}, \sigma_{2}), is a switching point for the differentiability of \mathcal{F}, if in any neighborhood U of \zeta_{0} there exist points \zeta_{1} < \zeta_{0} < \zeta_{2} such that
Type Ⅰ. at \zeta_{1} (2.1) holds while (2.2) does not hold and at \zeta_{2} (2.2) holds and (2.1) does not hold, or
Type Ⅱ. at \zeta_{1} (2.2) holds while (2.1) does not hold and at \zeta_{2} (2.1) holds and (2.2) does not hold.
Definition 2.7. ([23]) For \beta\in(0, 1] and let the left-sided \mathcal{GPF} -integral operator of order \vartheta of \mathcal{F} is defined as follows
\begin{eqnarray} \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}(\zeta) = \frac{1}{\beta^{\vartheta }\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}(\nu)d\nu,\quad\zeta > \sigma_{1}, \end{eqnarray} | (2.3) |
where \beta\in(0, 1] , \vartheta\in\mathbb{C}, Re(\vartheta) > 0 and \Gamma(.) is the Gamma function.
Definition 2.8. ([23]) For \beta\in(0, 1] and let the left-sided \mathcal{GPF} -derivative operator of order \vartheta of \mathcal{F} is defined as follows
\begin{eqnarray} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}(\zeta) = \frac{\mathcal{D}^{n,\beta}}{\beta^{n-\vartheta }\Gamma(n-\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{n-\vartheta-1}\mathcal{F}(\nu)d\nu, \end{eqnarray} | (2.4) |
where \beta\in(0, 1] , \vartheta\in\mathbb{C}, \, Re(\vartheta) > 0, \; n = [\vartheta]+1 and \mathcal{D}^{n, \beta} represents the nth -derivative with respect to proportionality index \beta.
Definition 2.9. ([23]) For \beta\in(0, 1] and let the left-sided \mathcal{GPF} -derivative in the sense of Caputo of order \vartheta of \mathcal{F} is defined as follows
\begin{eqnarray} \,^{c}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}(\zeta) = \frac{1}{\beta^{n-\vartheta }\Gamma(n-\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{n-\vartheta-1}\big(\mathcal{D}^{n,\beta}\mathcal{F}\big)(\nu)d\nu, \end{eqnarray} | (2.5) |
where \beta\in(0, 1] , \vartheta\in\mathbb{C}, \, Re(\vartheta) > 0 and n = [\vartheta]+1.
Let \Phi\in L([\sigma_{1}, \sigma_{2}], \mathfrak{E}), then the \mathcal{GPF} integral of order \vartheta of the fuzzy function \Phi is stated as:
\begin{eqnarray} \Phi_{\vartheta}^{\beta}(\zeta) = \Big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\Big)(\zeta) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\Phi(\nu)d\nu,\quad\zeta > \sigma_{1}. \end{eqnarray} | (2.6) |
Since [\Phi(\zeta)]^{\check{q}} = [\underline{\Phi}(\check{q}, \zeta), \bar{\Phi}(\check{q}, \zeta)] and 0 < \vartheta < 1, we can write the fuzzy \mathcal{GPF} -integral of the fuzzy mapping \Phi depend on lower and upper mappingss, that is,
\begin{eqnarray} \big[\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}{\Phi}\big)(\zeta)\big]^{\check{q}} = \big[\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\underline{\Phi}\big)(\check{q},\zeta),\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\bar{\Phi}\big)(\check{q},\zeta)\big],\quad \zeta\geq \sigma_{1}, \end{eqnarray} | (2.7) |
where
\begin{eqnarray} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\underline{\Phi}\big)(\check{q},\zeta) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\underline{\Phi}(\check{q},\nu)d\nu, \end{eqnarray} | (2.8) |
and
\begin{eqnarray} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\bar{\Phi}\big)(\check{q},\zeta) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\bar{\Phi}(\check{q},\nu)d\nu. \end{eqnarray} | (2.9) |
Definition 2.10. For n\in\mathbb{N}, order \vartheta and type \mathfrak{q} hold n-1 < \vartheta\leq n with 0\leq\mathfrak{q}\leq1. The left-sided fuzzy Hilfer-proportional \mathfrak{g}H -fractional derivative, with respect to \zeta having \beta\in(0, 1] of a function \zeta\in\mathcal{C}_{1-\gamma}^{\beta}[\sigma_{1}, \sigma_{2}], is stated as
\begin{eqnarray*} \Big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\Big)(\zeta) = \bigg(\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\Phi\big)\bigg)(\zeta), \end{eqnarray*} |
where \mathcal{D}^{\beta}\Phi(\nu) = (1-\beta)\Phi(\nu)+\beta \Phi^{\prime}(\nu) and if the \mathfrak{g}H -derivative \Phi_{(1-\vartheta), \beta}^{\prime}(\zeta) exists for \zeta\in[\sigma_{1}, \sigma_{2}], where
\begin{eqnarray*} \Phi_{(1-\vartheta)}^{\beta}(\zeta): = \Big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\vartheta),\beta}\Phi\Big)(\zeta) = \frac{1}{\beta^{1-\vartheta}\Gamma(1-\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta}{\Phi}(\nu)d\nu,\quad \zeta\geq \sigma_{1}. \end{eqnarray*} |
Definition 2.11. Let \Phi^{\prime}\in L([\sigma_{1}, \sigma_{2}], \mathfrak{E}) and the fractional generalized Hukuhara \mathcal{GPF} -derivative of fuzzy-valued function \Phi is stated as:
\begin{eqnarray} \big(\,_{\mathfrak{g}H}\mathcal{D}^{\vartheta,\beta}_{\sigma_{1}^{+}}\Phi\big)(\zeta) = \mathcal{I}_{\sigma_{1}^{+}}^{1-\vartheta,\beta}(\Phi^{\prime}_{\mathfrak{g}H})(\zeta) = \frac{1}{\beta^{1-\vartheta}\Gamma(1-\vartheta)}\int\limits_{\sigma_{1}}^{\vartheta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta}\Phi_{\mathfrak{g}H}^{\prime}(\nu)d\nu,\,\, \nu\in(\sigma_{1},\zeta). \end{eqnarray} | (2.10) |
Furthermore, we say that \Phi is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable at \zeta_{0} if
\begin{eqnarray} [(\,_{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta})]^{\check{q}}&& = \Bigg[\bigg[\frac{1}{\beta^{1-\vartheta}\Gamma(1-\vartheta)}\int\limits_{\sigma_{1}}^{\vartheta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta}\underline{\Phi}_{\mathfrak{g}H}^{\prime}(\nu)d\nu\bigg]^{\check{q}},\bigg[\frac{1}{\beta^{1-\vartheta}\Gamma(1-\vartheta)}\int\limits_{\sigma_{1}}^{\vartheta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta}\bar{\Phi}_{\mathfrak{g}H}^{\prime}(\nu)d\nu\bigg]^{\check{q}}\Bigg]\\&& = [(\,_{\mathfrak{g}H}\underline{\mathcal{D}}_{\sigma_{1}^{+}}^{\vartheta,\beta})(\check{q},\zeta),(\,_{\mathfrak{g}H}\bar{\mathcal{D}}_{\sigma_{1}^{+}}^{\vartheta,\beta})(\check{q},\zeta)] \end{eqnarray} | (2.11) |
and that \Phi is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable at \zeta_{0} if
\begin{eqnarray} [(\,_{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta})]^{\check{q}} = [(\,_{\mathfrak{g}H}\bar{\mathcal{D}}_{\sigma_{1}^{+}}^{\vartheta,\beta})(\check{q},\zeta),(\,_{\mathfrak{g}H}\underline{\mathcal{D}}_{\sigma_{1}^{+}}^{\vartheta,\beta})(\check{q},\zeta)]. \end{eqnarray} | (2.12) |
Definition 2.12. We say that a point \zeta_{0}\in(\sigma_{1}, \sigma_{2}), is a switching point for the differentiability of \mathcal{F}, if in any neighborhood U of \zeta_{0} there exist points \zeta_{1} < \zeta_{0} < \zeta_{2} such that
Type Ⅰ. at \zeta_{1} (2.11) holds while (2.12) does not hold and at \zeta_{2} (2.12) holds and (2.11) does not hold, or
Type Ⅱ. at \zeta_{1} (2.12) holds while (2.11) does not hold and at \zeta_{2} (2.11) holds and (2.12) does not hold.
Proposition 1. ([23]) Let \vartheta, \varrho\in\mathbb{C} such that Re(\vartheta) > 0 and Re(\varrho) > 0. Then for any \beta\in(0, 1], we have
\begin{eqnarray*} &&\Big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}e^{\frac{\beta-1}{\beta}}(s-\sigma_{1})^{\varrho-1}\Big)(\zeta) = \frac{\Gamma(\varrho)}{\beta^{\vartheta}\Gamma(\varrho+\vartheta)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\varrho+\vartheta-1},\\&&\Big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}e^{\frac{\beta-1}{\beta}}(s-\sigma_{1})^{\varrho-1}\Big)(\zeta) = \frac{\Gamma(\varrho)}{\beta^{\vartheta}\Gamma(\varrho-\vartheta)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\varrho-\vartheta-1},\\ &&\Big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}e^{\frac{\beta-1}{\beta}}(\sigma_{2}-s)^{\varrho-1}\Big)(\zeta) = \frac{\Gamma(\varrho)}{\beta^{\vartheta}\Gamma(\varrho+\vartheta)}e^{\frac{\beta-1}{\beta}(\sigma_{2}-s)}(\sigma_{2}-\zeta)^{\varrho+\vartheta-1},\\ &&\Big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}e^{\frac{\beta-1}{\beta}}(\sigma_{2}-s)^{\varrho-1}\Big)(\zeta) = \frac{\Gamma(\varrho)}{\beta^{\vartheta}\Gamma(\varrho-\vartheta)}e^{\frac{\beta-1}{\beta}(\sigma_{2}-s)}(\sigma_{2}-s)^{\varrho-\vartheta-1}. \end{eqnarray*} |
Lemma 2.13. ([24])For \beta\in(0, 1], \vartheta > 0, 0\leq\gamma < 1. If \Phi\in\mathcal{C}_{\gamma}[\sigma_{1}, \sigma_{2}] and \mathcal{I}_{\sigma_{1}^{+}}^{1-\vartheta}\Phi\in\mathcal{C}_{\gamma}^{1}[\sigma_{1}, \sigma_{2}], then
\begin{eqnarray*} \Big( \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\Big)(\zeta) = \Phi(\zeta)-\frac{e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\vartheta-1}}{\beta^{\vartheta-1}\Gamma(\vartheta)}\Big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\vartheta,\beta}\Phi\Big)(\sigma_{1}). \end{eqnarray*} |
Lemma 2.14. ([24]) Let \Phi\in L_{1}(\sigma_{1}, \sigma_{2}). If \mathcal{D}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta), \beta}\Phi exists on L_{1}(\sigma_{1}, \sigma_{2}), then
\begin{eqnarray*} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi = \mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\Phi. \end{eqnarray*} |
Lemma 2.15. Suppose there is a \mathfrak{d} -monotone fuzzy mapping \Phi\in\mathcal{AC}\big([\sigma_{1}, \sigma_{2}], \mathfrak{E}\big), where \big[\Phi(\zeta)\big]^{\check{q}} = \big[\underline{\Phi}(\check{q}, \zeta), \bar{\Phi}(\check{q}, \zeta)\big] for 0\leq\check{q}\leq1, \, \sigma_{1}\leq\zeta\leq \sigma_{2}, then for 0 < \vartheta < 1 and \beta\in(0, 1], we have
(i) \quad\big[\big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\Phi\big)(\zeta)\big]^{\check{q}} = \big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\underline{\Phi}(\check{q}, \zeta), \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\bar{\Phi}(\check{q}, \zeta)\big] for \zeta\in[\sigma_{1}, \sigma_{2}], if \Phi is \mathfrak{d} -increasing;
(ii)\quad \big[\big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\Phi\big)(\zeta)\big]^{\check{q}} = \big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\bar{\Phi}(\check{q}, \zeta), \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\underline{\Phi}(\check{q}, \zeta)\big] for \zeta\in[\sigma_{1}, \sigma_{2}], if \Phi is \mathfrak{d} -decreasing.
Proof. It is to be noted that if \Phi is \mathfrak{d} -increasing, then \big[\Phi^{\prime}(\zeta)\big]^{\check{q}} = \big[\frac{d}{d\zeta}\underline{\Phi}(\check{q}, \zeta), \frac{d}{d\zeta}\bar{\Phi}(\check{q}, \zeta)\big]. Taking into account Definition 2.10, we have
\begin{eqnarray*} \big[\big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta)\big]^{\check{q}}&& = \big[\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\underline{\Phi}\big)(\check{q},\zeta),\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\bar{\Phi}\big)(\check{q},\zeta)\big]\nonumber\\&& = \big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\underline{\Phi}(\check{q},\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\bar{\Phi}(\check{q},\zeta)\big]. \end{eqnarray*} |
If \Phi is \mathfrak{d} -decreasing, then \big[\Phi^{\prime}(\zeta)\big]^{\check{q}} = \big[\frac{d}{d\zeta}\bar{\Phi}(\check{q}, \zeta), \frac{d}{d\zeta}\underline{\Phi}(\check{q}, \zeta)\big], we have
\begin{eqnarray*} \big[\big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta)\big]^{\check{q}}&& = \big[\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\bar{\Phi}\big)(\check{q},\zeta),\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\underline{\Phi}\big)(\check{q},\zeta)\big]\nonumber\\&& = \big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\bar{\Phi}(\check{q},\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\underline{\Phi}(\check{q},\zeta)\big]. \end{eqnarray*} |
This completes the proof.
Lemma 2.16. For \beta\in(0, 1], \, \vartheta\in(0, 1). If \Phi\in\mathcal{AC}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) is a \mathfrak{d} -monotone fuzzy function. We take
z_{1}(\zeta): = \Big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\Big)(\zeta) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\Phi(\nu)d\nu, |
and
z_{1}^{(1-\vartheta),\beta}: = \Big(\mathcal{I}_{\sigma_{1}^{+}}^{(1-\vartheta),\beta}\Phi\Big)(\zeta) = \frac{1}{\beta^{1-\vartheta}\Gamma(1-\vartheta)}\int\limits_{\sigma_{1}}^{\vartheta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta}\Phi_{\mathfrak{g}H}^{\prime}(\nu)d\nu, |
is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}], then
\begin{eqnarray*} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta) = \Phi(\zeta)\ominus\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}, \end{eqnarray*} |
and
\begin{eqnarray*} \big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\big)(\zeta) = \Phi(\zeta). \end{eqnarray*} |
Proof. If z_{1}(\zeta) is \mathfrak{d} -increasing on [\sigma_{1}, \sigma_{2}] or z_{1}(\zeta) is \mathfrak{d} -decreasing on [\sigma_{1}, \sigma_{2}] and z_{1}^{(1-\vartheta), \beta}(\zeta) is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}].
Utilizing the Definitions 2.6, 2.10 and Lemma 2.13 with the initial condition (\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma, \beta}\Phi)(\sigma_{1}) = 0, we have
\begin{eqnarray} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta)&& = \bigg(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}^{\beta}\mathcal{I}_{\sigma_{1}^{+}}^{(1-\mathfrak{q})(1-\vartheta),\beta}\Phi\bigg)(\zeta)\\&& = \bigg(\mathcal{I}_{\sigma_{1}^{+}}^{\gamma,\beta}\mathcal{D}^{\beta}\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi\bigg)(\zeta)\\&& = \bigg(\mathcal{I}_{\sigma_{1}^{+}}^{\gamma,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\gamma,\beta}\Phi\bigg)(\zeta)\\&& = \Phi(\zeta)\ominus\frac{\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi}{\beta^{\gamma-1}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}. \end{eqnarray} | (2.13) |
Now considering Proposition 1, Lemma 2.13 and Lemma 2.14, we obtain
\begin{eqnarray*} \big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\big)(\zeta)&& = \bigg(\mathcal{I}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\mathfrak{q}(1-\vartheta),\beta}\Phi\bigg)(\zeta)\nonumber\\&& = \Phi(\zeta)\ominus\frac{\big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\mathfrak{q}(1-\vartheta),\beta}\Phi\big)(\sigma_{1})e^{\frac{\beta-1}{\beta}}(\zeta-\sigma_{1})}{\beta^{\mathfrak{q}(1-\vartheta)}\Gamma(\mathfrak{q}(1-\vartheta))}(\zeta-\sigma_{1})^{\mathfrak{q}(1-\vartheta)-1}\nonumber\\&& = \Phi(\zeta). \end{eqnarray*} |
On contrast, since \Phi\in\mathcal{AC}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) , there exists a constant \mathcal{K} such that \mathcal{K} = \sup\limits_{\zeta\in[\sigma_{1}, \sigma_{2}]}\bar{\mathcal{D}_{0}}[\Phi(\zeta), \hat{0}].
Then
\begin{eqnarray*} \bar{\mathcal{D}_{0}}[\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi(\zeta),\hat{0}] &&\leq\mathcal{K}\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}{e^{\frac{\beta-1}{\beta}(\zeta-\nu)}}{(\zeta-\nu)^{\vartheta-1}}d\nu\nonumber\\&&\leq\mathcal{K}\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}{\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert (\zeta-\nu)^{\vartheta-1}}d\nu\nonumber\\&& = \frac{\mathcal{K}}{\beta^{\vartheta}\Gamma(\vartheta+1)}(\zeta-\sigma_{1})^{\vartheta}, \end{eqnarray*} |
where we have used the fact \big\vert e^{\frac{\beta-1}{\beta}\zeta}\big\vert < 1 and \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta, \beta}\Phi(\zeta) = 0 and \zeta = \sigma_{1}.
This completes the proof.
Lemma 2.17. Let there be a continuous mapping \Phi:[\sigma_{1}, \sigma_{2}]\rightarrow \mathbb{R}^{+} on [\sigma_{1}, \sigma_{2}] and hold \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}\Phi(\zeta)\leq\mathcal{F}(\xi, \Phi(\xi)), \xi\geq\sigma_{1}, where \mathcal{F}\in\mathcal{C}([\sigma_{1}, \sigma_{1}]\times\mathbb{R}^{+}, \mathbb{R}^{+}). Assume that m(\zeta) = m(\zeta, \sigma_{1}, \xi_{0}) is the maximal solution of the IVP
\begin{eqnarray} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\xi(\zeta) = \mathcal{F}(\zeta,\xi),\quad\quad \big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\xi\big)(\sigma_{1}) = \xi_{0}\geq0, \end{eqnarray} | (2.14) |
on [\sigma_{1}, \sigma_{2}]. Then, if \Phi(\sigma_{1})\leq\xi_{0}, we have \Phi(\zeta)\leq m(\zeta), \zeta\in[\sigma_{1}, \sigma_{2}].
Proof. The proof is simple and can be derived as parallel to Theorem 2.2 in [53].
Lemma 2.18. Assume the IVP described as:
\begin{eqnarray} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi(\zeta) = \mathcal{F}(\zeta,\Phi(\zeta)),\quad\quad \big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi\big)(\sigma_{1}) = \Phi_{0} = 0,\quad\zeta\in[\sigma_{1},\sigma_{2}]. \end{eqnarray} | (2.15) |
Let \alpha > 0 be a given constant and \mathfrak{B}(\Phi_{0}, \alpha) = \big\{\Phi\in\mathbb{R}:\vert\Phi-\Phi_{0}\vert\leq\alpha\big\}. Assume that the real-valued functions \mathcal{F}:[\sigma_{1}, \sigma_{2}]\times[0, \alpha]\rightarrow \mathbb{R}^{+} satisfies the following assumptions:
(i) \mathcal{F}\in\mathcal{C}\big([\sigma_{1}, \sigma_{2}]\times[0, \alpha], \mathbb{R}^{+}\big), \, \mathcal{F}(\zeta, 0)\equiv0, \, 0\leq\mathcal{F}(\zeta, \Phi)\leq\mathcal{M}_{\mathcal{F}} for all (\zeta, \Phi)\in[\sigma_{1}, \sigma_{2}]\times[0, \alpha];
(ii) \mathcal{F}(\zeta, \Phi) is nondecreasing in \Phi for every \zeta\in[\sigma_{1}, \sigma_{2}]. Then the problem (2.15) has at least one solution defined on [\sigma_{1}, \sigma_{2}] and \Phi(\zeta)\in\mathfrak{B}(\Phi_{0}, \alpha).
Proof. The proof is simple and can be derived as parallel to Theorem 2.3 in [53].
In this investigation, we find the existence and uniqueness of solution to problem 1.3 by utilizing the successive approximation technique by considering the generalized Lipschitz condition of the right-hand side.
Lemma 3.1. For \gamma = \vartheta+\mathfrak{q}(1-\vartheta), \, \, \vartheta\in(0, 1), \mathfrak{q}\in[0, 1] with \beta\in(0, 1], and let there is a fuzzy function \mathcal{F}:(\sigma_{1}, \sigma_{2}]\times\mathfrak{E}\rightarrow \mathfrak{E} such that \zeta\rightarrow \mathcal{F}(\zeta, \Phi) belongs to \mathcal{C}_{\gamma}^{\beta}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) for any \Phi\in\mathfrak{E}. Then a \mathfrak{d} -monotone fuzzy function \Phi\in\mathcal{C}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) is a solution of IVP (1.3) if and only if \Phi satisfies the integral equation
\begin{eqnarray} &&\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\\&& = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu)\big)d\nu,\quad\zeta\in[\sigma_{1},\sigma_{2}],\; \; j = 1,2,...,m. \end{eqnarray} | (3.1) |
and the fuzzy function \zeta\rightarrow \mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma}\mathcal{F}(\zeta, \Phi) is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}].
Proof. Let \Phi\in\mathcal{C}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) be a \mathfrak{d} -monotone solution of (1.3), and considering z_{1}(\zeta): = \Phi(\zeta)\ominus_{\mathfrak{g}H}\big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma, \beta}\Phi\big)(\sigma_{1}), \zeta\in(\sigma_{1}, \sigma_{2}]. Since \Phi is \mathfrak{d} -monotone on [\sigma_{1}, \sigma_{2}], it follows that \zeta\rightarrow z_{1}(\zeta) is \mathfrak{d} -increasing on [\sigma_{1}, \sigma_{2}] (see [43]).
From (1.3) and Lemma 2.16, we have
\begin{eqnarray} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta) = \Phi(\zeta)\ominus\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1},\quad\forall \zeta\in[\sigma_{1},\sigma_{2}]. \end{eqnarray} | (3.2) |
Since \mathcal{F}(\zeta, \Phi)\in\mathcal{C}_{\gamma}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) for any \Phi\in\mathfrak{E}, and from (1.3), observes that
\begin{eqnarray} \big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\Phi\big)(\zeta) = \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}(\zeta,\Phi(\zeta)) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu)\big)d\nu,\quad\forall\zeta\in[\sigma_{1},\sigma_{2}]. \end{eqnarray} | (3.3) |
Additionally, since z_{1}(\zeta) is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}]. Also, we observe that \zeta\rightarrow \mathcal{F}^{\vartheta, \beta}(\zeta, \Phi) is also \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}].
Reluctantly, merging (3.2) and (3.3), we get the immediate consequence.
Further, suppose \Phi\in\mathcal{C}([\sigma_{1}, \sigma_{2}], \mathfrak{E}) be a \mathfrak{d} -monotone fuzzy function fulfills (3.1) and such that \zeta\rightarrow \mathcal{F}^{\vartheta, \beta}(\zeta, \Phi) is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}]. By the continuity of the fuzzy mapping \mathcal{F}, the fuzzy mapping \zeta\rightarrow \mathcal{F}^{\vartheta, \beta}(\zeta, \Phi) is continuous on (\sigma_{1}, \sigma_{2}] with \mathcal{F}^{\vartheta, \beta}(\sigma_{1}, \Phi(\sigma_{1})) = \lim\limits_{\zeta\rightarrow \sigma_{1}^{+}}\mathcal{F}^{\vartheta, \beta}(\zeta, \Phi) = 0. Then
\begin{eqnarray*} &&\Phi(\zeta) = \frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}\big(\zeta,\mathcal{\zeta}\big)\big)(\zeta),\nonumber\\&&\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi(\zeta) = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})+\big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\mathfrak{q}(1-\vartheta)}\mathcal{F}(\zeta,\Phi(\zeta))\big)(\zeta), \end{eqnarray*} |
and
\begin{eqnarray*} \mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi(0) = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j}). \end{eqnarray*} |
Moreover, since \zeta\rightarrow \mathcal{F}^{\vartheta, \beta}(\zeta, \Phi) is \mathfrak{d} -increasing on (\sigma_{1}, \sigma_{2}]. Applying, the operator \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta} on (3.1), yields
\begin{eqnarray*} &&\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\bigg(\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\bigg)\nonumber\\&& = \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\bigg(\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu)\big)d\nu\bigg)\nonumber\\&& = \mathcal{F}\big(\zeta,\Phi(\zeta)\big). \end{eqnarray*} |
This completes the proof.
In our next result, we use the following assumption. For a given constant \hbar > 0 , and let \mathfrak{B}(\Phi_{0}, \hbar) = \big\{\Phi\in\mathfrak{E}:\bar{\mathcal{D}_{0}}[\Phi, \Phi_{0}]\leq\hbar\big\}.
Theorem 3.2. Let \mathcal{F}\in\mathcal{C}\big([\sigma_{1}, \sigma_{2}]\times\mathfrak{B}(\Phi_{0}, \hbar), \mathfrak{E}\big) and suppose that the subsequent assumptions hold:
(i) there exists a positive constant \mathcal{M}_{\mathcal{F}} such that \bar{\mathcal{D}_{0}}[\mathcal{F}(\zeta, z_{1}), \hat{0}]\leq\mathcal{M}_{\mathcal{F}}, for all (\zeta, z_{1})\in[\sigma_{1}, \sigma_{2}]\times\mathfrak{B}(\Phi_{0}, \hbar) ;
(ii) for every \zeta\in[\sigma_{1}, \sigma_{2}] and every z_{1}, \omega\in\mathfrak{B}(\Phi_{0}, \hbar),
\begin{eqnarray} \bar{\mathcal{D}_{0}}\big[\mathcal{F}(\zeta,z_{1}),\mathcal{F}(\zeta,\omega)\big]\leq\mathfrak{g}(\zeta,\bar{\mathcal{D}_{0}}[z_{1},\omega]), \end{eqnarray} | (3.4) |
where \mathfrak{g}(\zeta, .)\in\mathcal{C}\big([\sigma_{1}, \sigma_{2}]\times[0, \beta], \mathbb{R}^{+}\big) satisfies the assumption in Lemma 2.18 given that problem (2.15) has only the solution \phi(\zeta)\equiv0 on [\sigma_{1}, \sigma_{2}]. Then the subsequent successive approximations given by \Phi^{0}(\zeta) = \Phi_{0} and for n = 1, 2, ...,
\begin{eqnarray*} &&\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big)d\nu, \end{eqnarray*} |
converges consistently to a fixed point of problem (1.3) on certain interval [\sigma_{1}, \mathcal{T}] for some \mathcal{T}\in(\sigma_{1}, \sigma_{2}] given that the mapping \zeta\rightarrow \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta, \beta}\mathcal{F}(\zeta, \Phi^{n}(\zeta)) is \mathfrak{d} -increasing on [\sigma_{1}, \mathcal{T}].
Proof. Take \sigma_{1} < \zeta^{*} such that \zeta^{*}\leq\big[\frac{\beta^{\vartheta}\hbar.\Gamma(1+\vartheta)}{\mathcal{M}}+\sigma_{1}\big]^{\frac{1}{\vartheta}}, where \mathcal{M} = \max\big\{\mathcal{M}_{\mathfrak{g}}, \mathcal{M}_{\mathcal{F}}\big\} and put \mathcal{T}: = \min\{\zeta^{*}, \sigma_{2}\}. Let \mathbb{S} be a set of continuous fuzzy functions \Phi such that \omega(\sigma_{1}) = \Phi_{0} and \omega(\zeta)\in\mathfrak{B}(\Phi_{0}, \hbar) for all \zeta\in[\sigma_{1}, \mathcal{T}]. Further, we suppose the sequence of continuous fuzzy function \{\Phi^{n}\}_{n = 0}^{\infty} given by \Phi^{0}(\zeta) = \Phi_{0}, \, \forall \zeta\in[\sigma_{1}, \mathcal{T}] and for n = 1, 2, ..,
\begin{eqnarray} &&\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\\&& = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big)d\nu. \end{eqnarray} | (3.5) |
Firstly, we show that \Phi^{n}(\zeta)\in\mathcal{C}([\sigma_{1}, \mathcal{T}], \mathfrak{B}(\Phi_{0}, \hbar)) . For n\geq1 and for any \zeta_{1}, \zeta_{2}\in[\sigma_{1}, \mathcal{T}] with \zeta_{1} < \zeta_{2}, we have
\begin{eqnarray*} &&\bar{\mathcal{D}_{0}}\Bigg(\Phi^{n}(\zeta_{1})\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1},\Phi^{n}(\zeta_{2})\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\Bigg)\nonumber\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta_{1}}\Big[e^{\frac{\beta-1}{\beta}(\zeta_{1}-\nu)}(\zeta_{1}-\nu)^{\vartheta-1}-e^{\frac{\beta-1}{\beta}(\zeta_{2}-\nu)}(\zeta_{2}-\nu)^{\vartheta-1}\Big]\bar{\mathcal{D}_{0}}\big[\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big),\hat{0}\big]d\nu\nonumber\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{1}}^{\zeta_{2}}e^{\frac{\beta-1}{\beta}(\zeta_{2}-\nu)}(\zeta_{2}-\nu)^{\vartheta-1}\bar{\mathcal{D}_{0}}\big[\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big),\hat{0}\big]d\nu. \end{eqnarray*} |
Using the fact that \vert e^{\frac{\beta-1}{\beta}\zeta}\vert < 1, then, on the right-hand side from the last inequality, the subsequent integral becomes \frac{1}{\beta^{\vartheta}\Gamma(1+\vartheta)}(\zeta_{2}-\zeta_{1})^{\vartheta}. Therefore, with the similar assumption as we did above, the first integral reduces to \frac{1}{\beta^{\vartheta}\Gamma(1+\vartheta)}\big[(\zeta_{1}-\sigma_{1})^{\vartheta}-(\zeta_{2}-\sigma_{1})^{\vartheta}+(\zeta_{2}-\zeta_{1})^{\vartheta}\big]. Thus, we conclude
\begin{eqnarray*} \bar{\mathcal{D}_{0}}\big[\Phi^{n}\big((\zeta_{1}),\Phi^{n}(\zeta_{2})\big)\big]&&\leq\frac{\mathcal{M}_{\mathcal{F}}}{\beta^{\vartheta}\Gamma(1+\vartheta)}\big[(\zeta_{1}-\sigma_{1})^{\vartheta}-(\zeta_{2}-\sigma_{1})^{\vartheta}+2(\zeta_{2}-\zeta_{1})^{\vartheta}\big]\nonumber\\&&\leq\frac{2\mathcal{M}_{\mathcal{F}}}{\beta^{\vartheta}\Gamma(1+\vartheta)}(\zeta_{2}-\zeta_{1})^{\vartheta}. \end{eqnarray*} |
In the limiting case as \zeta_{1}\rightarrow \zeta_{2}, then the last expression of the above inequality tends to 0, which shows \Phi^{n} is a continuous function on [\sigma_{1}, \mathcal{T}] for all n\geq1.
Moreover, it follows that \Phi^{n}\in\mathfrak{B}(\Phi_{0}, \hbar) for all n\geq0, \, \zeta\in[\sigma_{1}, \mathcal{T}] if and only if \Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\in\mathfrak{B}(0, \hbar) for all \zeta\in[\sigma_{1}, \mathcal{T}] and for all n\geq0.
Also, if we assume that \Phi^{n-1}(\zeta)\in\mathbb{S} for all \zeta\in[\sigma_{1}, \mathcal{T}], \, n\geq2, then
\begin{eqnarray*} &&\bar{\mathcal{D}_{0}}\Big[\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}, \hat{0}\Big]\nonumber\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\bar{\mathcal{D}_{0}}\big[\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big),\hat{0}\big]d\nu\nonumber\\&& = \frac{\mathcal{M}_{\mathcal{F}}(\zeta-\sigma_{1})^{\vartheta}}{\beta^{\vartheta}\Gamma(1+\vartheta)}\leq\hbar. \end{eqnarray*} |
It follows that \Phi^{n}(\zeta)\in\mathbb{S}, \, \forall\in[\sigma_{1}, \mathcal{T}].
Henceforth, by mathematical induction, we have \Phi^{n}(\zeta)\in\mathbb{S}, \, \, \forall \zeta\in[\sigma_{1}, \mathcal{T}] and \forall\, n\geq1.
Further, we show that the sequence \Phi^{n}(\zeta) converges uniformly to a continuous function \Phi\in\mathcal{C}([\sigma_{1}, \mathcal{T}], \mathfrak{B}(\Phi_{0}, \hbar)). By assertion (ii) and mathematical induction, we have for \zeta\in[\sigma_{1}, \mathcal{T}]
\begin{eqnarray} &&\bar{\mathcal{D}_{0}}\bigg[\Phi^{n+1}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1},\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\bigg]\\&&\leq \phi^{n}(\zeta),\quad\quad n = 0,1,2,..., \end{eqnarray} | (3.6) |
where \phi^{n}(\zeta) is defined as follows:
\begin{eqnarray} \phi^{n}(\zeta) = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathfrak{g}\big(\nu,\phi^{n-1}(\nu)\big)d\nu, \end{eqnarray} | (3.7) |
where we have used the fact that \vert e^{\frac{\beta-1}{\beta}\zeta}\vert < 1 and \phi^{0}(\zeta) = \frac{\mathcal{M}(\zeta-\sigma_{1})^{\vartheta}}{\beta^{{\vartheta}}\Gamma({\vartheta}+1)}. Thus, we have, for \zeta\in[\sigma_{1}, \mathcal{T}] and for n = 0, 1, 2, ...,
\begin{eqnarray*} &&\bar{\mathcal{D}_{0}}\big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n+1}(\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n}(\zeta)\big] \nonumber\\&&\leq\bar{\mathcal{D}_{0}}\big[\mathcal{F}(\zeta,\Phi^{n}(\zeta)),\mathcal{F}(\zeta,\Phi^{n-1}(\zeta))\big]\nonumber\\&&\leq\mathfrak{g}\big(\zeta,\bar{\mathcal{D}_{0}}\big[\Phi^{n}(\zeta),\Phi^{n-1}(\zeta)\big]\big)\nonumber\\&&\leq\mathfrak{g}\big(\zeta,\phi^{n-1}(\zeta)\big). \end{eqnarray*} |
Let n\leq m and \zeta\in[\sigma_{1}, \mathcal{T}], then one obtains
\begin{eqnarray*} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\bar{\mathcal{D}_{0}}\big[\Phi^{n}(\zeta),\Phi^{m}(\zeta)\big]&&\leq\bar{\mathcal{D}_{0}}\big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n}(\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{m}(\zeta)\big]\nonumber\\&&\leq\bar{\mathcal{D}_{0}}\big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n}(\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n+1}(\zeta)\big]+\bar{\mathcal{D}_{0}}\big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{n+1}(\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{m+1}(\zeta)\big]\nonumber\\&&\quad+\bar{\mathcal{D}_{0}}\big[\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{m+1}(\zeta),\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi^{m}(\zeta)\big]\nonumber\\&&\leq2\mathfrak{g}(\zeta,\phi^{n-1}(\zeta))+\mathfrak{g}\big(\zeta,\bar{\mathcal{D}_{0}}[\Phi^{n}(\zeta),\Phi^{m}(\zeta)]\big). \end{eqnarray*} |
From (ii), we observe that the solution \phi(\zeta) = 0 is a unique solution of problem (2.15) and \mathfrak{g}(., \phi^{n-1}):[\sigma_{1}, \mathcal{T}]\rightarrow [0, \mathcal{M}_{\mathfrak{g}}] uniformly converges to 0 , for every \epsilon > 0, there exists a natural number n_{0} such that
\begin{eqnarray*} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\bar{\mathcal{D}_{0}}\big[\Phi^{n}(\zeta),\Phi^{m}(\zeta)\big]\leq\mathfrak{g}\big(\zeta,\bar{\mathcal{D}_{0}}[\Phi^{n}(\zeta),\Phi^{m}(\zeta)]\big)+\epsilon,\quad\quad for\,n_{0}\leq n\leq m. \end{eqnarray*} |
Using the fact that \bar{\mathcal{D}_{0}}\big[\Phi^{n}(\sigma_{1}), \Phi^{m}(\sigma_{1})\big] = 0 < \epsilon and by using Lemma 2.17, we have for \zeta\in[\sigma_{1}, \mathcal{T}]
\begin{eqnarray} \bar{\mathcal{D}_{0}}\big[\Phi^{n}(\zeta),\Phi^{m}(\zeta)\big]\leq \delta_{\epsilon}(\zeta),\quad n_{0}\leq n\leq m, \end{eqnarray} | (3.8) |
where \delta_{\epsilon}(\zeta) is the maximal solution to the following IVP:
\begin{eqnarray*} \big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\delta_{\epsilon}\big)(\zeta) = \mathfrak{g}(\zeta,\delta_{\epsilon}(\zeta))+\epsilon,\quad\quad \big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma}\delta_{\epsilon}\big) = \epsilon. \end{eqnarray*} |
Taking into account Lemma 2.17, we deduce that [\phi_{\epsilon}(., \omega)] converges uniformly to the maximal solution \phi(\zeta) \equiv0 of (2.15) on [\sigma_{1}, \mathcal{T}] as \epsilon\rightarrow 0.
Therefore, in view of (3.8), we can obtain n_{0}\in\mathbb{N} is large enough such that, for n_{0} < n, m,
\begin{eqnarray} &&\sup\limits_{\zeta\in[\sigma_{1},\mathcal{T}]}\bar{\mathcal{D}_{0}}\bigg[\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1},\Phi^{m}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\bigg]\\&&\leq\epsilon. \end{eqnarray} | (3.9) |
Since (\mathfrak{E}, \bar{\mathcal{D}_{0}}) is a complete metric space and (3.9) holds, thus \big\{\Phi^{n}(\zeta)\} converges uniformly to \Phi\in\mathcal{C}([\sigma_{1}, \sigma_{2}], \mathfrak{B}(\Phi_{0}, \hbar)). Hence
\begin{eqnarray} \Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}&& = \lim\limits_{n\rightarrow \infty}\bigg(\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi^{n-1}(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\bigg)\\&& = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big)d\nu. \end{eqnarray} | (3.10) |
Because of Lemma 3.1, the function \Phi(\zeta) is the solution to (1.3) on [\sigma_{1}, \mathcal{T}].
In order to find the unique solution, assume that \Psi:[\sigma_{1}, \mathcal{T}]\rightarrow \mathfrak{E} is another solution of problem (1.3) on [\sigma_{1}, \mathcal{T}]. We denote \kappa(\zeta) = \bar{\mathcal{D}_{0}}[\Phi(\zeta), \Psi(\zeta)]. Then \kappa(\sigma_{1}) = 0 and for every \zeta\in[\sigma_{1}, \mathcal{T}], we have
\begin{eqnarray} \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q},\beta}\kappa(\zeta)\leq\bar{\mathcal{D}_{0}}\big[\mathcal{F}(\zeta,\Phi(\zeta)),\mathcal{F}(\zeta,\Psi(\zeta))\big]\leq\mathfrak{g}(\zeta,\kappa(\zeta)). \end{eqnarray} | (3.11) |
Further, using the comaprison Lemma 2.17, we get \kappa(\zeta)\leq m(\zeta), where m is a maximal solution of the IVP \mathcal{D}_{\sigma_{1}^{+}}^{\vartheta, \mathfrak{q}, \beta}m(\zeta)\leq\mathfrak{g}(\zeta, m(\zeta)), \, \big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma}m\big)(\sigma_{1}) = 0. By asseration (ii), we have m(\zeta) = 0 and hence \Phi(\zeta) = \Psi(\zeta), \, \forall\in[\sigma_{1}, \mathcal{T}].
This completes the proof.
Corollary 1. For \beta\in(0, 1] and let \mathcal{C}([\sigma_{1}, \sigma_{2}], \mathfrak{E}). Assume that there exist positive constants \mathcal{L}, \mathcal{M}_{\mathcal{F}} such that, for every z_{1}, \omega\in\mathfrak{E},
\begin{eqnarray*} \bar{\mathcal{D}_{0}}\big[\mathcal{F}(\zeta,z_{1}),\mathcal{F}(\zeta,\omega)\big]\leq\mathcal{L}\bar{\mathcal{D}_{0}}[z_{1},\omega],\quad\quad \bar{\mathcal{D}_{0}}\big[\mathcal{F}(\zeta,z_{1}),\hat{0}\big]\leq\mathcal{M}_{\mathcal{F}}. \end{eqnarray*} |
Then the subsequent successive approximations given by \Phi^{0}(\zeta) = \Phi_{0} and for n = 1, 2, ..
\begin{eqnarray*} \Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\Phi_{0} = \frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi^{n-1}(\nu)\big)d\nu, \end{eqnarray*} |
converges consistently to a fixed point of problem (1.3) on [\sigma_{1}, \mathcal{T}] for certain \mathcal{T}\in(\sigma_{1}, \sigma_{2}] given that the mapping \zeta\rightarrow \mathcal{I}_{\sigma_{1}^{+}}^{\vartheta, \beta}\mathcal{F}(\zeta, \Phi^{n}(\zeta)) is \mathfrak{d} -increasing on [\sigma_{1}, \mathcal{T}].
Example 3.3. For \beta\in(0, 1], \, \gamma = \vartheta+\mathfrak{q}(1-\vartheta), \, \vartheta\in(0, 1), \, \mathfrak{q}\in[0, 1] and \delta\in\mathbb{R}. Assume that the linear fuzzy \mathcal{GPF} - FDE under Hilfer- \mathcal{GPF} -derivative and moreover, the subsequent assumptions hold:
\begin{eqnarray} \begin{cases} \big(\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\mathfrak{q}}\Phi\big)(\zeta) = \delta\Phi(\zeta)+\eta(\zeta),\quad\quad\quad\quad\quad\quad\zeta\in(\sigma_{1},\sigma_{2}],\\ \big(\mathcal{I}_{\sigma_{1}^{+}}^{1-\gamma,\beta}\Phi\big)(\sigma_{1}) = \Phi_{0} = \sum\limits_{j = 1}^{m}\mathcal{R}_{j}\Phi(\zeta_{j}),\quad\quad \gamma = \vartheta+\mathfrak{q}(1-\vartheta). \end{cases} \end{eqnarray} | (3.12) |
Applying Lemma 3.1, we have
\begin{eqnarray*} &&\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \delta\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\Phi(\nu)d\nu+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\sigma_{1}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\eta(\nu)d\nu,\quad\zeta\in[\sigma_{1},\sigma_{2}]\nonumber\\&& = \delta\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\big)(\zeta)+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta), \end{eqnarray*} |
where \eta\in\mathcal{C}((\sigma_{1}, \sigma_{2}], \mathfrak{E}) and furthermore, assuming the diameter on the right part of the aforementioned equation is increasing. Observing \mathcal{F}(\zeta, \Phi): = \delta\Phi+\eta fulfill the suppositions of Corollary 1.
In order to find the analytical view of (3.12), we utilized the technique of successive approximation. Putting \Phi^{0}(\zeta) = \Phi_{0} and
\begin{eqnarray*} &&\Phi^{n}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \delta\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi^{n-1}\big)(\zeta)+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta),\quad n = 1,2,... \end{eqnarray*} |
Letting n = 1, \; \delta > 0, assuming there is a \mathfrak{d} -increasing mapping \Phi, then we have
\begin{eqnarray*} &&\Phi^{1}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \delta\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\frac{(\zeta-\sigma_{1})^{\vartheta}}{\beta^{\vartheta}\Gamma(\vartheta+1)}+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta). \end{eqnarray*} |
In contrast, if we consider \delta < 0 and \Phi is \mathfrak{d} -decreasing, then we have
\begin{eqnarray*} &&(-1)\bigg(\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\ominus_{\mathfrak{g}H}\Phi^{1}(\zeta)\bigg)\nonumber\\&& = \delta\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\frac{(\zeta-\sigma_{1})^{\vartheta}}{\beta^{\vartheta}\Gamma(\vartheta+1)}+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta). \end{eqnarray*} |
For n = 2 , we have
\begin{eqnarray*} &&\Phi^{2}(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\bigg[\frac{\delta(\zeta-\sigma_{1})^{\vartheta}}{\beta^{\vartheta}\Gamma(\vartheta+1)}+\frac{\delta^{2}(\zeta-\sigma_{1})^{2\vartheta}}{\beta^{2\vartheta}\Gamma(2\vartheta+1)}\bigg]+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta)+\big(\mathcal{I}_{\sigma_{1}^{+}}^{2\vartheta,\beta}\eta\big)(\zeta), \end{eqnarray*} |
if \delta > 0 and there is \mathfrak{d} -increasing mapping \Phi , we have
\begin{eqnarray*} &&(-1)\bigg(\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\ominus_{\mathfrak{g}H}\Phi^{2}(\zeta)\bigg)\nonumber\\&& = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\bigg[\frac{\delta(\zeta-\sigma_{1})^{\vartheta}}{\beta^{\vartheta}\Gamma(\vartheta+1)}+\frac{\delta^{2}(\zeta-\sigma_{1})^{2\vartheta}}{\beta^{2\vartheta}\Gamma(2\vartheta+1)}\bigg]+\big(\mathcal{I}_{\sigma_{1}^{+}}^{\vartheta,\beta}\eta\big)(\zeta)+\big(\mathcal{I}_{\sigma_{1}^{+}}^{2\vartheta,\beta}\eta\big)(\zeta), \end{eqnarray*} |
and there is \delta < 0, \, and\, \mathfrak{d} -increasing mapping \Phi. So, continuing inductively and in the limiting case, when n\rightarrow \infty, we attain the solution
\begin{eqnarray*} &&\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\sum\limits_{l = 1}^{\infty}\frac{\delta^{l}(\zeta-\sigma_{1})^{l\vartheta}}{\beta^{l\vartheta}\Gamma(l\vartheta+1)}+\int\limits_{\sigma_{1}}^{\zeta}\sum\limits_{l = 1}^{\infty}\frac{\delta^{l-1}(\zeta-\sigma_{1})^{l\vartheta}-1}{\beta^{l\vartheta-1}\Gamma(l\vartheta)}\eta(\nu)d\nu\nonumber\\&& = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\sum\limits_{l = 1}^{\infty}\frac{\delta^{l}(\zeta-\sigma_{1})^{l\vartheta}}{\beta^{l\vartheta}\Gamma(l\vartheta+1)}+\int\limits_{\sigma_{1}}^{\zeta}\sum\limits_{l = 0}^{\infty}\frac{\delta^{l}(\zeta-\sigma_{1})^{l\vartheta+(\vartheta-1)}}{\beta^{l\vartheta}+(\vartheta-1)\Gamma(l\vartheta+\vartheta)}\eta(\nu)d\nu\nonumber\\&& = \sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})\sum\limits_{l = 1}^{\infty}\frac{\delta^{l}(\zeta-\sigma_{1})^{l\vartheta}}{\beta^{l\vartheta}\Gamma(l\vartheta+1)}+\frac{1}{\beta^{\vartheta-1}}\int\limits_{\sigma_{1}}^{\zeta}(\zeta-\sigma_{1})^{\vartheta-1}\sum\limits_{l = 0}^{\infty}\frac{\delta^{l}(\zeta-\sigma_{1})^{l\vartheta}}{\beta^{l\vartheta}\Gamma(l\vartheta+\vartheta)}\eta(\nu)d\nu, \end{eqnarray*} |
for every \delta > 0 and \Phi is \mathfrak{d} -increasing, or \delta < 0 and \Phi is \mathfrak{d} -decreasing, accordingly. Therefore, by means of Mittag-Leffler function \mathcal{E}_{\vartheta, \mathfrak{q}}(\Phi) = \sum\limits_{l = 1}^{\infty}\frac{\Phi^{\kappa}}{\Gamma(l\vartheta+\mathfrak{q})}, \, \vartheta, \mathfrak{q} > 0, the solution of problem (3.12) is expressed by
\begin{eqnarray*} &&\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = {\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}\mathcal{E}_{\vartheta,1}\Big({\delta(\zeta-\sigma_{1})^{\vartheta}}\Big)+\frac{1}{\beta^{\vartheta-1}}\int\limits_{\sigma_{1}}^{\zeta}(\zeta-\sigma_{1})^{\vartheta-1}\mathcal{E}_{\vartheta,\vartheta}\Big({\delta(\zeta-\sigma_{1})^{\vartheta}}\Big)\eta(\nu)d\nu, \end{eqnarray*} |
for every of \delta > 0 and \Phi is \mathfrak{d} -increasing. Alternately, if \delta < 0 and \Phi is \mathfrak{d} -decreasing, then we get the solution of problem (3.12)
\begin{eqnarray*} &&\Phi(\zeta)\ominus_{\mathfrak{g}H}\frac{\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}{\beta^{\gamma}\Gamma(\gamma)}e^{\frac{\beta-1}{\beta}(\zeta-\sigma_{1})}(\zeta-\sigma_{1})^{\gamma-1}\nonumber\\&& = {\sum\limits_{j = 1}^{m}R_{j}\Phi(\zeta_{j})}\mathcal{E}_{\vartheta,1}\Big({\delta(\zeta-\sigma_{1})^{\vartheta}}\Big)\ominus(-1)\frac{1}{\beta^{\vartheta-1}}\int\limits_{\sigma_{1}}^{\zeta}(\zeta-\sigma_{1})^{\vartheta-1}\mathcal{E}_{\vartheta,\vartheta}\Big({\delta(\zeta-\sigma_{1})^{\vartheta}}\Big)\eta(\nu)d\nu. \end{eqnarray*} |
Consider IVP
\begin{equation} \begin{cases} \big(\,_{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi\big)(\zeta) = \mathcal{F}\big(\zeta,\Phi(\zeta),\mathcal{H}_{1}\Phi(\zeta),\mathcal{H}_{2}\Phi(\zeta)\big),\quad\quad \zeta\in[\zeta_{0},\mathcal{T}]\\\Phi(\zeta_{0}) = \Phi_{0}\in\mathfrak{E}, \end{cases} \end{equation} | (4.1) |
where \beta\in(0, 1] and \vartheta\in(0, 1) is a real number and the operation _{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta} denote the \mathcal{GPF} derivative of order \vartheta, \mathcal{F}:[\zeta_{0}, \mathcal{T}]\times\mathfrak{E}\times\mathfrak{E}\times\mathfrak{E}\rightarrow \mathfrak{E} is continuous in \zeta which fulfills certain supposition that will be determined later, and
\begin{eqnarray} \mathcal{H}_{1}\Phi(\zeta) = \int\limits_{\zeta_{0}}^{\zeta}\mathcal{H}_{1}(\zeta,s)\Phi(s)ds,\quad\quad\quad\mathcal{H}_{2}\Phi(\zeta) = \int\limits_{\zeta_{0}}^{\mathcal{T}}\mathcal{H}_{2}(\zeta,s)\Phi(s)ds, \end{eqnarray} | (4.2) |
with \mathcal{H}_{1}, \mathcal{H}_{2}:[\zeta_{0}, \mathcal{T}]\times[\zeta_{0}, \mathcal{T}]\rightarrow \mathbb{R} such that
\begin{eqnarray*} \mathcal{H}_{1}^{*} = \sup\limits_{\zeta\in[\zeta_{0},\mathcal{T}]}\int\limits_{\zeta_{0}}^{\zeta}\vert\mathcal{H}_{1}(\zeta,s)\vert ds,\quad\quad\quad\mathcal{H}_{2}^{*} = \sup\limits_{\zeta\in[\zeta_{0},\mathcal{T}]}\int\limits_{\zeta_{0}}^{\mathcal{T}}\vert\mathcal{H}_{2}(\zeta,s)\vert ds. \end{eqnarray*} |
Now, we investigate the existence and uniqueness of the solution of problem (4.1). To establish the main consequences, we require the following necessary results.
Theorem 4.1. Let \mathcal{F}:[\sigma_{1}, \sigma_{2}]\rightarrow \mathfrak{E} be a fuzzy-valued function on [\sigma_{1}, \sigma_{2}]. Then
(i) \mathcal{F} is [(i)-\mathfrak{g}H] -differentiable at c\in[\sigma_{1}, \sigma_{2}] iff \mathcal{F} is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable at c.
(ii) \mathcal{F} is [(ii)-\mathfrak{g}H] -differentiable at c\in[\sigma_{1}, \sigma_{2}] iff \mathcal{F} is \, ^{GPF}[(ii)-\mathfrak{g}H] -differentiable at c.
Proof. In view of Definition 2.18 and Definition 2.11, the proof is straightforward.
Lemma 4.2. ([44]) Let there be a fuzzy valued mapping \mathcal{F}:[\zeta_{0}, \mathcal{T}]\rightarrow \mathfrak{E} such that \mathcal{F}^{\prime}_{\mathfrak{g}H}\in\mathfrak{E}\cap\chi_{c}^{r}(\sigma_{1}, \sigma_{2}), then
\begin{eqnarray} \mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\big(\,_{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\mathcal{F}\big)(\zeta) = \mathcal{F}(\zeta)\ominus_{\mathfrak{g}H}\mathcal{F}(\zeta_{0}). \end{eqnarray} | (4.3) |
Lemma 4.3. The IVP (4.1) is analogous to subsequent equation
\begin{eqnarray} \Phi(\zeta) = \Phi_{0}+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu, \end{eqnarray} | (4.4) |
if \Phi(\zeta) be \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable,
\begin{eqnarray} \Phi(\zeta) = \Phi_{0}\ominus\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu, \end{eqnarray} | (4.5) |
if \Phi(\zeta) be \, ^{GPF}[(ii)-\mathfrak{g}H] -differentiable, and
\begin{eqnarray} \Phi(\zeta) = \begin{cases} \Phi_{0}+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu,\quad\zeta\in[\sigma_{1},\sigma_{3}],\\ \Phi_{0}\ominus\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu,\quad\zeta\in[\sigma_{3},\sigma_{2}], \end{cases} \end{eqnarray} | (4.6) |
if there exists a point \sigma_{3}\in(\sigma_{1}, \sigma_{2}) such that \Phi(\zeta) is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable on [\sigma_{1}, \sigma_{3}] and \, ^{GPF}[(ii)-\mathfrak{g}H] -differentiable on [\sigma_{3}, \sigma_{2}] and \mathcal{F}(\sigma_{3}, \Phi(\sigma_{3}, \Phi(\sigma_{3}), \mathcal{H}_{1}\Phi(\sigma_{3}))\in\mathbb{R}.
Proof. By means of the integral operator (2.6) on both sides of (4.1), yields
\begin{eqnarray} \mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\big(\,_{\mathfrak{g}H}\mathcal{D}_{\sigma_{1}^{+}}^{\vartheta,\beta}\Phi(\zeta)\big) = \mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\big(\mathcal{F}(\zeta,\Phi(\zeta),\mathcal{H}_{1}\Phi(\zeta),\mathcal{H}_{2}\Phi(\zeta)\big). \end{eqnarray} | (4.7) |
Utilizing Lemma 4.2 and Definition 2.6, we gat
\begin{eqnarray} \Phi(\zeta)\ominus_{\mathfrak{g}H}\Phi_{0} = \frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu. \end{eqnarray} | (4.8) |
In view of Defnition 2.17 and Theorem 4.1, if \Phi(\zeta) be \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable,
\begin{eqnarray} \Phi(\zeta) = \Phi_{0}+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu \end{eqnarray} | (4.9) |
and if \Phi(\zeta) be \, ^{GPF}[(ii)-\mathfrak{g}H] -differentiable
\begin{eqnarray} \Phi(\zeta) = \Phi_{0}\ominus\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu. \end{eqnarray} | (4.10) |
In addition, when we have a switchpoint \sigma_{3}\in(\sigma_{1}, \sigma_{2}) of type (I) the \, ^{GPF}[\mathfrak{g}H] -differentiability changes from type (I) to type (II) at \zeta = \sigma_{3}. Then by (4.9) and (4.10) and Definition 2.12, The proof is easy to comprehend.
Also, we proceed with the following assumptions:
({\mathbb{A}_{1}}). \mathcal{F}:[\zeta_{0}, \mathcal{T}]\times\mathfrak{E}\times\mathfrak{E}\times\mathfrak{E}\rightarrow \mathfrak{E} is continuous and there exist positive real functions \mathcal{L}_{1}, \mathcal{L}_{2}, \mathcal{L}_{3} such that
\begin{eqnarray*} &&\bar{\mathcal{D}_{0}}\Big(\mathcal{F}(\zeta,\Phi(\zeta),\mathcal{H}_{1}\Phi(\zeta),\mathcal{H}_{2}\Phi(\zeta)),\mathcal{F}(\zeta,\Psi(\zeta),\mathcal{H}_{1}\Psi(\zeta),\mathcal{H}_{2}\Psi(\zeta))\Big)\nonumber\\&&\leq\mathcal{L}_{1}(\zeta)\bar{\mathcal{D}_{0}}(\Phi,\Psi)+\mathcal{L}_{2}(\zeta)\bar{\mathcal{D}_{0}}(\mathcal{H}_{1}\Phi,\mathcal{H}_{1}\Psi)+\mathcal{L}_{3}(\zeta)\bar{\mathcal{D}_{0}}(\mathcal{H}_{2}\Phi,\mathcal{H}_{2}\Psi). \end{eqnarray*} |
({\mathbb{A}_{2}}). There exist a number \epsilon such that \delta\leq\epsilon < 1, \, \zeta\in[\zeta_{0}, \mathcal{T}]
\delta = \mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{P}(1+\mathcal{H}_{1}^{*}+\mathcal{H}_{2}^{*}) |
and
\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{P} = \sup\limits_{\zeta\in[0,\mathcal{T}]}\big\{\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{1},\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{2},\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{3}\big\}. |
Theorem 4.4. Let \mathcal{F}:[\zeta_{0}, \mathcal{T}]\times\mathfrak{E}\times\mathfrak{E}\times\mathfrak{E}\rightarrow \mathfrak{E} be a bounded continuous functions and holds (\mathbb{A}_{1}). Then the IVP (4.1) has a unique solution which is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiable on [\zeta_{0}, \mathcal{T}], given that \delta < 1, where \delta is given in (\mathbb{A}_{2}).
Proof. Assuming \Phi(\zeta) is \, ^{GPF}[(i)-\mathfrak{g}H] -differentiability and \Phi_{0}\in\mathfrak{E} be fixed. Propose a mapping \mathfrak{F}:\mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E})\rightarrow \mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}) by
\begin{eqnarray} \big(\mathfrak{F}\Phi\big)(\zeta) = \Phi_{0}+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu,\quad for\,all\,\zeta\in[\zeta_{0},\mathcal{T}]. \end{eqnarray} | (4.11) |
Next we prove that \mathfrak{F} is contraction. For \Phi, \Psi\in\mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}) by considering of (\mathbb{A}_{1}) and by distance properties (2.3), one has
\begin{eqnarray} &&\bar{\mathcal{D}_{0}}\big(\mathfrak{F}\Phi(\zeta),\mathfrak{F}\Psi(\zeta)\big)\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert \bar{\mathcal{D}_{0}}\big(\mathcal{F}(\zeta,\Phi(\zeta),\mathcal{H}_{1}\Phi(\zeta),\mathcal{H}_{2}\Phi(\zeta)),\mathcal{F}(\zeta,\Psi(\zeta),\mathcal{H}_{1}\Psi(\zeta),\mathcal{H}_{2}\Psi(\zeta))\Big)d\nu\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\big[\mathcal{L}_{1}\bar{\mathcal{D}_{0}}(\Phi,\Psi)+\mathcal{L}_{2}\bar{\mathcal{D}_{0}}(\mathcal{H}_{1}\Phi,\mathcal{H}_{1}\Psi)+\mathcal{L}_{3}\bar{\mathcal{D}_{0}}(\mathcal{H}_{2}\Phi,\mathcal{H}_{2}\Psi) \big]d\nu\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{1}\bar{\mathcal{D}_{0}}(\Phi,\Psi)d\nu+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{2}\bar{\mathcal{D}_{0}}(\mathcal{H}_{1}\Phi,\mathcal{H}_{1}\Psi)d\nu\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{3}\bar{\mathcal{D}_{0}}(\mathcal{H}_{2}\Phi,\mathcal{H}_{2}\Psi)d\nu. \end{eqnarray} | (4.12) |
Now, we find that
\begin{eqnarray} &&\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{2}\bar{\mathcal{D}_{0}}(\mathcal{H}_{1}\Phi,\mathcal{H}_{1}\Psi)d\nu\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\Big(\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{2}\bar{\mathcal{D}_{0}}(\Phi,\Psi)\int\limits_{\zeta_{0}}^{\nu}\vert\mathcal{H}_{1}(\nu,x)\vert dx \Big)d\nu\\&&\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{2}\mathcal{H}_{1}^{*}.\bar{\mathcal{D}_{0}}(\Phi,\Psi). \end{eqnarray} | (4.13) |
Analogously,
\begin{eqnarray} &&\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{3}\bar{\mathcal{D}_{0}}(\mathcal{H}_{2}\Phi,\mathcal{H}_{2}\Psi)d\nu\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{3}\mathcal{H}_{1}^{*}.\bar{\mathcal{D}_{0}}(\Phi,\Psi),\\ &&\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{1}\bar{\mathcal{D}_{0}}(\Phi,\Psi)d\nu = \mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{1}\bar{\mathcal{D}_{0}}(\Phi,\Psi). \end{eqnarray} | (4.14) |
Then we have
\begin{eqnarray} \bar{\mathcal{D}_{0}}\big(\mathfrak{F}\Phi,\mathfrak{F}\Psi\big)&&\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{1}\bar{\mathcal{D}_{0}}(\Phi,\Psi)+\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{2}\mathcal{H}_{1}^{*}.\bar{\mathcal{D}_{0}}(\Phi,\Psi)+\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{L}_{3}\mathcal{H}_{2}^{*}.\bar{\mathcal{D}_{0}}(\Phi,\Psi)\\&&\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{P}(1+\mathcal{H}_{1}^{*}+\mathcal{H}_{2}^{*})\bar{\mathcal{D}_{0}}(\Phi,\Psi)\\&& < \bar{\mathcal{D}_{0}}(\Phi,\Psi). \end{eqnarray} | (4.15) |
Consequently, \mathfrak{F} is a contraction mapping on \mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}) having a fixed point \mathfrak{F}\Phi(\zeta) = \Phi(\zeta). Henceforth, the IVP (4.1) has unique solution.
Theorem 4.5. For \beta\in(0, 1] and let \mathcal{F}:[\zeta_{0}, \mathcal{T}]\times\mathfrak{E}\times\mathfrak{E}\times\mathfrak{E}\rightarrow \mathfrak{E} be a bounded continuous functions and satisfies (\mathbb{A}_{1}). Let the sequence \Phi_{n}:[\zeta_{0}, \mathcal{T}]\rightarrow \mathfrak{E} is given by
\begin{eqnarray} \Phi_{n+1}(\zeta)&& = \Phi_{0}\ominus\frac{-1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi_{n}(\nu),\mathcal{H}_{1}\Phi_{n}(\nu),\mathcal{H}_{2}\Phi_{n}(\nu)\big)d\nu,\quad\\ \Phi_{0}(\zeta)&& = \Phi_{0}, \end{eqnarray} | (4.16) |
is described for any n\in\mathbb{N}. Then the sequence \{\Phi_{n}\} converges to fixed point of problem (4.1) which is \, ^{GPF}[(ii)-\mathfrak{g}H] -differentiable on [\zeta_{0}, \mathcal{T}], given that \delta < 1, where \delta is defined in (\mathbb{A}_{2}).
Proof. We now prove that the sequence \{\Phi_{n}\} , given in (4.16), is a Cauchy sequence in \mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}). To do just that, we'll require
\begin{eqnarray} \bar{\mathcal{D}_{0}}(\Phi_{1},\Phi_{0})&& = \bar{\mathcal{D}_{0}}\bigg(\Phi_{0}\ominus\frac{-1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi_{0}(\nu),\mathcal{H}_{1}\Phi_{0}(\nu),\mathcal{H}_{2}\Phi_{0}(\nu)\big)d\nu,\Phi_{0}\bigg)\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\bar{\mathcal{D}_{0}}\Big(\mathcal{F}\big(\nu,\Phi_{0}(\nu),\mathcal{H}_{1}\Phi_{0}(\nu),\mathcal{H}_{2}\Phi_{0}(\nu)\big), \hat{0}\Big)d\nu\\&&\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{M}, \end{eqnarray} | (4.17) |
where \mathcal{M} = \sup_{\zeta\in[\zeta_{0}, \mathcal{T}]}\bar{\mathcal{D}_{0}}\big(\mathcal{F}(\zeta, \Phi, \mathcal{H}_{1}\Phi, \mathcal{H}_{2}\Phi), \hat{0}\big).
Since \mathcal{F} is Lipschitz continuous, In view of Definition (2.3), we show that
\begin{eqnarray} &&\bar{\mathcal{D}_{0}}(\Phi_{n+1},\Phi_{n})\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert \bar{\mathcal{D}_{0}}\big(\mathcal{F}\big(\nu,\Phi_{n}(\nu),\mathcal{H}_{1}\Phi_{n}(\nu),\mathcal{H}_{2}\Phi_{n}(\nu)\big),\mathcal{F}\big(\nu,\Phi_{n-1}(\nu),\mathcal{H}_{1}\Phi_{n-1}(\nu),\mathcal{H}_{2}\Phi_{n-1}(\nu)\big)\Big)d\nu\\&&\leq\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{1}.\bar{\mathcal{D}_{0}}\big(\Phi_{n},\Phi_{n-1}\big)d\nu\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{2}.\bar{\mathcal{D}_{0}}\big(\mathcal{H}_{1}\Phi_{n},\mathcal{H}_{1}\Phi_{n-1}\big)d\nu\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{3}.\bar{\mathcal{D}_{0}}\big(\mathcal{H}_{2}\Phi_{n},\mathcal{H}_{2}\Phi_{n-1}\big)d\nu\\&&\leq\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{P}(1+\mathcal{H}_{1}^{*}+\mathcal{H}_{2}^{*})\bar{\mathcal{D}_{0}}(\Phi_{n},\Phi_{n-1})\leq\delta \bar{\mathcal{D}_{0}}(\Phi_{n},\Phi_{n-1})\leq\delta^{n}\bar{\mathcal{D}_{0}}(\Phi_{1},\Phi_{0})\leq\delta^{n}\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{M}. \end{eqnarray} | (4.18) |
Since \delta < 1 promises that the sequence \{\Phi_{n}\} is a Cauchy sequence in \mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}). Consequently, there exist \Phi\in\mathcal{C}([\zeta_{0}, \mathcal{T}], \mathfrak{E}) such that \{\Phi_{n}\} converges to \Phi. Thus, we need to illustrate that \Phi is a solution of the problem (4.1).
\begin{eqnarray} &&\bar{\mathcal{D}}_{0}\bigg(\Phi(\zeta)+\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu, \Phi_{0}\bigg)\\&& = \bar{\mathcal{D}}_{0}\bigg(\Phi(\zeta)+\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu, \Phi_{n+1}\\&&\quad+\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi_{n}(\nu),\mathcal{H}_{1}\Phi_{n}(\nu),\mathcal{H}_{2}\Phi_{n}(\nu)\big)d\nu\bigg)\\&&\leq \bar{\mathcal{D}_{0}}\big(\Phi(\zeta),\Phi_{n+1}\big)+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{1}.\bar{\mathcal{D}_{0}}\big(\Phi(\nu),\Phi_{n}\big)d\nu\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{2}.\bar{\mathcal{D}_{0}}\big(\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{1}\Phi_{n}\big)d\nu\\&&\quad+\frac{1}{\beta^{\vartheta}\Gamma(\vartheta)}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert\mathcal{L}_{3}.\bar{\mathcal{D}_{0}}\big(\mathcal{H}_{2}\Phi(\nu),\mathcal{H}_{2}\Phi_{n}\big)d\nu\\&&\leq \bar{\mathcal{D}_{0}}\big(\Phi(\zeta),\Phi_{n+1}\big)+\mathcal{I}_{\zeta_{0}}^{\vartheta,\beta}\mathcal{P}(1+\mathcal{H}_{1}^{*}+\mathcal{H}_{2}^{*})\bar{\mathcal{D}_{0}}(\Phi(\zeta),\Phi_{n}). \end{eqnarray} | (4.19) |
In the limiting case, when n\rightarrow \infty. Thus we have
\begin{eqnarray} \Phi(\zeta)+\frac{-1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}e^{\frac{\beta-1}{\beta}(\zeta-\nu)}(\zeta-\nu)^{\vartheta-1}\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu)\big)d\nu = \Phi_{0}. \end{eqnarray} | (4.20) |
By Lemma 4.3, we prove that \Phi is a solution of the problem (4.1). In order to prove the uniqness of \Phi(\zeta), let \Psi(\zeta) be another solution of problem (4.1) on [\zeta_{0}, \mathcal{T}]. Utilizing Lemma 4.3, gets
\begin{eqnarray*} \bar{\mathcal{D}_{0}}(\Phi,\Psi)\leq\frac{1}{\beta^{\vartheta}\Gamma(\mathfrak{q})}\int\limits_{\zeta_{0}}^{\zeta}\big\vert e^{\frac{\beta-1}{\beta}(\zeta-\nu)}\big\vert\big\vert(\zeta-\nu)^{\vartheta-1}\big\vert \bar{\mathcal{D}}_{0}\bigg(\mathcal{F}\big(\nu,\Phi(\nu),\mathcal{H}_{1}\Phi(\nu),\mathcal{H}_{2}\Phi(\nu),\mathcal{F}\big(\nu,\Psi(\nu),\mathcal{H}_{1}\Psi(\nu),\mathcal{H}_{2}\Psi(\nu)\big)\bigg)d\nu. \end{eqnarray*} |
Analogously, by employing the distance properties \bar{\mathcal{D}}_{0} and Lipschitiz continuity of \mathcal{F}, consequently, we deduce that (1-\delta)\bar{\mathcal{D}_{0}}(\Phi, \Psi)\leq0, since \delta < 1, we have \Phi(\zeta) = \Psi(\zeta) for all \zeta\in[\zeta_{0}, \mathcal{T}]. Hence, the proof is completed.
Example 4.6. Suppose the Cauchy problem by means of differential operator (2.4)
\begin{eqnarray} \mathcal{D}_{z}^{\vartheta,\beta}\Phi(z) = \mathcal{F}(z,\Phi(z)), \end{eqnarray} | (4.21) |
where \mathcal{F}(z, \Phi(z)) is analytic in \Phi and \Phi(z) is analytic in the unit disk. Therefore, \mathcal{F} can be written as
\begin{eqnarray*} \mathcal{F}(z,\Phi) = \varphi \Phi(z). \end{eqnarray*} |
Consider \mathcal{Z} = z^{\vartheta}. Then the solution can be formulated as follows:
\begin{eqnarray} \Phi(\mathcal{Z}) = \sum\limits_{j = 0}^{\infty}\Phi_{j}\mathcal{Z}^{j}, \end{eqnarray} | (4.22) |
where \Phi_{j} are constants. Putting (4.22) in (4.21), yields
\begin{eqnarray*} \frac{\partial}{\partial z}\sum\limits_{j = 0}^{\infty}\Upsilon_{\vartheta,\beta,j}\Phi_{j}\mathcal{Z}^{j}-\varphi\sum\limits_{j = 0}^{\infty}\Phi_{j}\mathcal{Z}^{j} = 0. \end{eqnarray*} |
Since
\begin{eqnarray*} \Upsilon_{\vartheta,\beta,j} = \frac{\beta^{\vartheta}\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}{j\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}, \end{eqnarray*} |
then the simple computations gives the expression
\begin{eqnarray*} \frac{\beta^{\vartheta}\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}{\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}\Phi_{j}-\varphi\Phi_{j-1} = 0. \end{eqnarray*} |
Consequently, we get
\begin{eqnarray*} \Phi_{j} = \Big(\frac{\varphi}{\beta^{\vartheta}}\Big)^{j}\frac{\Gamma\big(\frac{(j-1) {\vartheta}}{\beta}+1-\vartheta\big)\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}{\Gamma\big(\frac{(j-1)\vartheta}{\beta}+1\big)\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}. \end{eqnarray*} |
Therefore, we have the subsequent solution
\begin{eqnarray*} \Phi(\mathcal{Z}) = \sum\limits_{j = 0}^{\infty}\Big(\frac{\varphi}{\beta^{\vartheta}}\Big)^{j}\frac{\Gamma\big(\frac{(j-1) {\vartheta}}{\beta}+1-\vartheta\big)\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}{\Gamma\big(\frac{(j-1)\vartheta}{\beta}+1\big)\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}\mathcal{Z}^{j}, \end{eqnarray*} |
or equivalently
\begin{eqnarray*} \Phi(\mathcal{Z}) = \sum\limits_{j = 0}^{\infty}\Big(\frac{\varphi}{\beta^{\vartheta}}\Big)^{j}\frac{\Gamma(j+1)\Gamma\big(\frac{(j-1) {\vartheta}}{\beta}+1-\vartheta\big)\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}{\Gamma\big(\frac{(j-1)\vartheta}{\beta}+1\big)\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}\frac{\mathcal{Z}^{j}}{j!}, \end{eqnarray*} |
where \varphi is assumed to be arbitrary constant, we take
\varphi: = \beta^{\vartheta}. |
Therefore, for appropriate \vartheta, we have
\begin{eqnarray*} \Phi(\mathcal{Z})&& = \sum\limits_{j = 0}^{\infty}\Big(\frac{\varphi}{\beta^{\vartheta}}\Big)^{j}\frac{\Gamma(j+1)\Gamma\big(\frac{(j-1) {\vartheta}}{\beta}+1-\vartheta\big)\Gamma\big(\frac{j\vartheta}{\beta}+1-\vartheta\big)}{\Gamma\big(\frac{(j-1)\vartheta}{\beta}+1\big)\Gamma\big(\frac{j\vartheta}{\beta}+1\big)}\frac{\mathcal{Z}^{j}}{j!}\nonumber\\&& = \,_{3}\Psi_{2}\begin{bmatrix} (1,1),\Big(1-\vartheta-\frac{\vartheta}{\beta},\frac{\vartheta}{\beta}\Big),\Big(1-\vartheta,\frac{\vartheta}{\beta}\Big);\\\qquad\qquad\qquad\quad\quad\quad\quad\quad\qquad\qquad\qquad\qquad\mathcal{Z}\\\Big(1-\frac{\vartheta}{\beta},\frac{\vartheta}{\beta},\Big),\Big(1,\frac{\vartheta}{\beta}\Big); \end{bmatrix}\nonumber\\&& = \,_{3}\Psi_{2}\begin{bmatrix} (1,1),\Big(1-\vartheta-\frac{\vartheta}{\beta},\frac{\vartheta}{\beta}\Big),\Big(1-\vartheta,\frac{\vartheta}{\beta}\Big);\\\qquad\qquad\qquad\quad\quad\quad\quad\quad\qquad\qquad\qquad\qquad z^{\vartheta\beta}\\\Big(1-\frac{\vartheta}{\beta},\frac{\vartheta}{\beta},\Big),\Big(1,\frac{\vartheta}{\beta}\Big); \end{bmatrix}, \end{eqnarray*} |
where \vert z\vert < 1.
The present investigation deal with an IVP for \mathcal{GPF} fuzzy FDEs and we employ a new scheme of successive approximations under generalized Lipschitz condition to obtain the existence and uniqueness consequences of the solution to the specified problem. Furthermore, another method to discover exact solutions of \mathcal{GPF} fuzzy FDEs by utilizing the solutions of integer order differential equations is considered. Additionally, the existence consequences for \mathcal{FVFIDE}s under \mathcal{GPF} - \mathcal{HD} with fuzzy initial conditions are proposed. Also, the uniqueness of the so-called integrodifferential equations is verified. Meanwhile, we derived the equivalent integral forms of the original fuzzy \mathcal{FVFIDE}s whichis utilized to examine the convergence of these arrangements of conditions. Two examples enlightened the efficacy and preciseness of the fractional-order \mathcal{HD} and the other one presents the exact solution by means of the Fox-Wright function. For forthcoming mechanisms, we will relate the numerical strategies for the estimated solution of nonlinear fuzzy FDEs.
The authors would like to express their sincere thanks to the support of Taif University Researchers Supporting Project Number (TURSP-2020/217), Taif University, Taif, Saudi Arabia.
The authors declare that they have no competing interests.
[1] |
Abdullah SM, Siddiqua S, Siddiquee MSH, et al. (2017) Modeling and forecasting exchange rate volatility in Bangladesh using GARCH models: a comparison based on normal and Student's t-error distribution. Financ Innov 3: 18. https://doi.org/10.1186/s40854-017-0071-z doi: 10.1186/s40854-017-0071-z
![]() |
[2] |
Abounoori A, Elmi ZM, Nademi Y (2016) Forecasting Tehran stock exchange volatility; Markov switching GARCH approach. Phys A 445: 264–282. https://doi.org/10.1016/j.physa.2015.10.024 doi: 10.1016/j.physa.2015.10.024
![]() |
[3] |
Atabani Adi A (2019) Modeling exchange rate return volatility of RMB/USD using GARCH family models. J Chinese Econ Bus Stud 17: 169–187. https://doi.org/10.1080/14765284.2019.1600933 doi: 10.1080/14765284.2019.1600933
![]() |
[4] |
Baillie R, Bollerslev T (2002) The message in daily exchange rates: a conditional-variance tale. J Bus Econ Stat 20–1: 60–68. https://doi.org/10.1198/073500102753410390 doi: 10.1198/073500102753410390
![]() |
[5] |
Barndorff-Nielsen OE (1997) Normal inverse Gaussian distributions and stochastic volatility modeling. Scand J Stat 24: 1–13. https://doi.org/10.1111/1467-9469.t01-1-00045 doi: 10.1111/1467-9469.t01-1-00045
![]() |
[6] |
Barndorff-Nielsen OE, Levendorskii SZ (2001) Feller processes of normal inverse Gaussian type. Quant Financ 1: 318–331. https://doi.org/10.1088/1469-7688/1/3/303 doi: 10.1088/1469-7688/1/3/303
![]() |
[7] | Barndorff-Nielsen OE, Schmiegel J (2008) Time change, volatility, and turbulence. Math Control Theory Financ, 29–53. |
[8] | Barndorff-Nielsen OE, Shephard N (2001) Normal Modified Stable Processes. Oxford: Department of Economics, Oxford University Press. https://doi.org/10.1007/978-1-4612-0197-7 |
[9] | Black F (1976) Studies of Stock Price Volatility Changes. Proceedings of the Business and Economics Section of the American Statistical Association, 177–181. |
[10] |
Bollerslev T (1986) Generalized autoregressive conditional heteroskedasticity. J Econ 31–3: 307–327. https://doi.org/10.1016/0304-4076(86)90063-1 doi: 10.1016/0304-4076(86)90063-1
![]() |
[11] |
Calzolari G, Halbleib R, Parrini L (2014) Estimating GARCH-type models with symmetric stable innovations: Indirect inference versus maximum likelihood. Comput Stat Data Anal 76: 158–171. https://doi.org/10.1016/j.csda.2013.07.028 doi: 10.1016/j.csda.2013.07.028
![]() |
[12] |
Cont R (2001) Empirical properties of asset returns: Stylized facts and statistical issues. Quantit Financ 1: 1–14. https://doi.org/10.1080/713665670 doi: 10.1080/713665670
![]() |
[13] |
Danielsson J, De Vries CG (2000) Value-at-risk and extreme returns. Ann Econ Stat 60: 239–270. https://doi.org/10.2307/20076262 doi: 10.2307/20076262
![]() |
[14] |
Engle RF (1982) Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation. Econometrica 50–4: 987–1007. https://doi.org/10.2307/1912773 doi: 10.2307/1912773
![]() |
[15] |
Feng L, Shi Y (2017) Fractionally integrated GARCH model with tempered stable distribution: a simulation study. J Appl Stat 44–16: 2837–2857. https://doi.org/10.1080/02664763.2016.1266310 doi: 10.1080/02664763.2016.1266310
![]() |
[16] |
Fiorani F, Luciano E, Semeraro P (2010) Single and joint default in a structural model with purely discountinuous asset prices. Quantit Financ 10: 249–263. https://doi.org/10.1080/14697680902991965 doi: 10.1080/14697680902991965
![]() |
[17] | Hamilton JD (1994) Time Series Analysis. Princeton University Press, Princeton. https://doi.org/10.1515/9780691218632 |
[18] |
Hirsa A, Madan DB (2003) Pricing american options under variance gamma. J Comput Financ 7: 63–80. https://doi.org/10.21314/JCF.2003.112 doi: 10.21314/JCF.2003.112
![]() |
[19] |
Ho KY, Shi Y, Zhang Z (2013) How does news sentiment impact asset volatility? Evidence from long memory and regime-switching approaches. N Am J Econ Financ 26: 436–456. https://doi.org/10.1016/j.najef.2013.02.015 doi: 10.1016/j.najef.2013.02.015
![]() |
[20] |
Kim YS, Rachev ST, Bianchi L, et al. (2010) Tempered stable and tempered infinitely divisible GARCH models. J Bank Financ 34: 2096–2109. https://doi.org/10.1016/j.jbankfin.2010.01.015 doi: 10.1016/j.jbankfin.2010.01.015
![]() |
[21] |
Letac G, Seshadri V (2005) Exponential stopping and drifted stable processes. Stat Prob Lett 72: 137–143. https://doi.org/10.1016/j.spl.2004.12.007 doi: 10.1016/j.spl.2004.12.007
![]() |
[22] | Lopez JA (2001) Evaluating the predictive accuracy of volatility models. J Forecast 20: 87–109. |
[23] |
Louati M, Masmoudi A, Mselmi F (2020) The normal tempered stable regression model. Commun Stat-Theor M 49: 500–512. https://doi.org/10.1080/03610926.2018.1554121 doi: 10.1080/03610926.2018.1554121
![]() |
[24] |
Madan DB, Carr P, Chang E (1998) The variance gamma process and option pricing. Eur Financ Rev 2: 79–105. https://doi.org/10.1023/A:1009703431535 doi: 10.1023/A:1009703431535
![]() |
[25] |
Madan DB, Seneta E (1990) The Variance Gamma (V.G.) model for share market returns. J Bus 63: 511–524. https://doi.org/10.1086/296519 doi: 10.1086/296519
![]() |
[26] |
Marinelli C, Rachev ST, Roll R (2001) Subordinated exchange rate models: Evidence for heavy tailed distributions and long-range dependence. Math Comput Model 34: 955–1001. https://doi.org/10.1016/S0895-7177(01)00113-3 doi: 10.1016/S0895-7177(01)00113-3
![]() |
[27] |
Nelson DB (1991) Conditional heteroskedasticity in asset returns: A new approach. Econometrica 59: 347–370. https://doi.org/10.2307/2938260 doi: 10.2307/2938260
![]() |
[28] | Nolan J (2007) Stable Distributions: Models for Heavy Tailed Data. Brikhauser, Boston. |
[29] |
Podobnik B, Horvatic D, Petersen AM, et al. (2009) Cross-correlations between volume change and price change. Proc Natl Aca 106: 22079–22084. https://doi.org/10.1073/pnas.0911983106 doi: 10.1073/pnas.0911983106
![]() |
[30] |
Rodriguez MJ, Ruiz E (2012) Revisiting Several Popular GARCH Models with Leverage Effect: Differences and Similarities. J Financ Econometrics 10: 637–668. https://doi.org/10.1093/jjfinec/nbs003 doi: 10.1093/jjfinec/nbs003
![]() |
[31] |
Shi Y, Feng L (2016) A discussion on the innovation distribution of the Markov regime-switching GARCH model. Econ Model 53: 278–288. https://doi.org/10.1016/j.econmod.2015.11.018 doi: 10.1016/j.econmod.2015.11.018
![]() |
[32] | Tserakh U, Trusz M (2019) GARCH(1, 1) models with stable residuals. Stud Inf Syst Inf Tech 22: 47–57. |
[33] |
Zhu D, Galbraith JW (2011) Modeling and forecasting expected shortfall with the generalized asymmetric student-t and asymmetric exponential power distributions. J Emp Fin 18: 765–778. https://doi.org/10.1016/j.jempfin.2011.05.006 doi: 10.1016/j.jempfin.2011.05.006
![]() |
1. | Maysaa Al-Qurashi, Saima Rashid, Fahd Jarad, Madeeha Tahir, Abdullah M. Alsharif, New computations for the two-mode version of the fractional Zakharov-Kuznetsov model in plasma fluid by means of the Shehu decomposition method, 2022, 7, 2473-6988, 2044, 10.3934/math.2022117 | |
2. | Saima Rashid, Rehana Ashraf, Zakia Hammouch, New generalized fuzzy transform computations for solving fractional partial differential equations arising in oceanography, 2023, 8, 24680133, 55, 10.1016/j.joes.2021.11.004 | |
3. | Saima Rashid, Rehana Ashraf, Ahmet Ocak Akdemir, Manar A. Alqudah, Thabet Abdeljawad, Mohamed S. Mohamed, Analytic Fuzzy Formulation of a Time-Fractional Fornberg–Whitham Model with Power and Mittag–Leffler Kernels, 2021, 5, 2504-3110, 113, 10.3390/fractalfract5030113 | |
4. | Saima Rashid, Rehana Ashraf, Fatimah S. Bayones, A Novel Treatment of Fuzzy Fractional Swift–Hohenberg Equation for a Hybrid Transform within the Fractional Derivative Operator, 2021, 5, 2504-3110, 209, 10.3390/fractalfract5040209 | |
5. | Saima Rashid, Sobia Sultana, Bushra Kanwal, Fahd Jarad, Aasma Khalid, Fuzzy fractional estimates of Swift-Hohenberg model obtained using the Atangana-Baleanu fractional derivative operator, 2022, 7, 2473-6988, 16067, 10.3934/math.2022880 | |
6. | Shuang-Shuang Zhou, Saima Rashid, Asia Rauf, Khadija Tul Kubra, Abdullah M. Alsharif, Initial boundary value problems for a multi-term time fractional diffusion equation with generalized fractional derivatives in time, 2021, 6, 2473-6988, 12114, 10.3934/math.2021703 | |
7. | Manar A. Alqudah, Rehana Ashraf, Saima Rashid, Jagdev Singh, Zakia Hammouch, Thabet Abdeljawad, Novel Numerical Investigations of Fuzzy Cauchy Reaction–Diffusion Models via Generalized Fuzzy Fractional Derivative Operators, 2021, 5, 2504-3110, 151, 10.3390/fractalfract5040151 | |
8. | Ravichandran VIVEK, Kangarajan K., Dvivek VİVEK, Elsayed ELSAYED, Dynamics and Stability of \Xi-Hilfer Fractional Fuzzy Differential Equations with Impulses, 2023, 6, 2651-4001, 115, 10.33434/cams.1257750 | |
9. | Saima Rashid, Fahd Jarad, Hind Alamri, New insights for the fuzzy fractional partial differential equations pertaining to Katugampola generalized Hukuhara differentiability in the frame of Caputo operator and fixed point technique, 2024, 15, 20904479, 102782, 10.1016/j.asej.2024.102782 |