Abstract

The aim of this paper is to derive a solution for a generalized free electron laser equation in terms of the incomplete Mittag-Leffler function and in terms of the incomplete Wright function..
Keywords: Incomplete Mittag-Leffler function, Incomplete Wright function, Caputo Derivative.

1 Introduction

Fractional Calculus is one of the most important tools for theoretical and experimental physics. Fractional calculus is of great scientific significance as it bestows a unified approach to various physical problems and yields fundamental exploratory ideas.In view of the effect of fractional calculus, it has been used by many researchers [2, 3, 4, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16] in the study of free electron lasers. Dattoli et al. [6, 9] described the unsaturated behavior of the free electron laser in the first-order integro-differential equation of Volterra type as

$$D_{\mu}h(\mu)=-i\pi g_{0}\int_{0}^{\mu}\psi h(\mu-\psi)e^{i\nu\psi}d\psi,~{}~{}~{}~{}~{}0\leq\mu\leq 1.$$ (1.1)

where $\mu$ is a dimensionless time variable, $g_{0}$ is a positive constant called the small-signal gain and the detuning parameter is the constant $\nu$. The function $h(\mu)$ is the complex-field amplitude which is assumed to be dimensionless satisfying the initial condition $h(0)=1$. The exact closed-form solution of (1.1) valid in the region of practical interest and suitable for numerical computation is given in [8].
The fractional integral of an arbitrary order is the generalization of the n-th order integral$(n\in N)$. The most fundamental definition of the fractional integral of arbitrary order $a,$ due to Riemann and Liouville [1], is given by

$$(R^{a}_{0,x}f)(x)=(I^{a}_{0,x}f)(x)=(D^{-a}_{0,x}f)(x)=\frac{1}{\Gamma(a)}\int_{0}^{x}(x-t)^{a-1}f(t)dt$$ (1.2)

$$=\frac{x^{a}}{\Gamma(a)}\int_{0}^{1}(1-u)^{a-1}f(xu)du,~{}~{}~{}~{}(\Re(a)>0).$$ (1.3)

and

$$(D^{a}_{0,x}f)(x)=D^{n}_{0,x}(D^{a-n}_{0,x}f)(x)=D^{n}_{0,x}(I^{n-a}_{0,x}f)(x).$$ (1.4)

Laplace transform of the fractional derivative

$$L\{D^{a}_{t};s\}=s^{a}F(s)-\sum_{r=1}^{n}s^{r-1}D^{a-r}_{t}f(t)|_{t}=0,~{}~{}~{}~{}(r-1<a\leq r;r\in N).$$ (1.5)

The Caputo derivatives of the casual function $f(t)$ (that is $f(t)=0$ for $t<0$) with $a>0$ was defined by Caputo [14] in the form

$$_{0}^{c}D^{a}_{t}f(t)=(_{0}^{c}D^{a}_{t}f)(t)=\frac{d^{a}}{dx^{a}}f(t)=\frac{1}{\Gamma(m-a)}\int_{0}^{\tau}\frac{f^{(m)}(u)}{(t-u)^{a-m+1}}du,$$ (1.6)

where $m-1<a<m,m\in N$ and $f^{(m)}(t)$ denote the usual derivative of $f(t)$ of order $m\in N_{0}$.
The Laplace transform of the Caputo derivative in the form

$$L\{\frac{d^{a}}{dt^{a}}f(t);s\}=s^{a}F(s)-\sum_{r=0}^{m-1}s^{a-r-1}f^{(r)}(0);m-1<a\leq m;m\in N,$$ (1.7)

where $F(s)$ is the Laplace transform of $f(t)$

$$L\{f(t)\}=F(s)=\int_{0}^{\infty}e^{-st}f(t)dt,$$ (1.8)

where $\Re(s)>0$, which may be symbolically written as

$$F(s)=L\{f(t);s\}\text{(or)}f(t)=L^{-1}\{F(s);t\}.$$

The homogenous form of equation (1.1) is given by Boyadjiev et al. [12], which is as follows

$$D^{a}_{\mu}h(\mu)=\lambda\int_{0}^{\mu}\psi h(\mu-\psi)e^{i\nu\psi}d\psi+be^{i\mu},~{}~{}~{}~{}~{}0\leq\mu\leq 1,$$ (1.9)

where $b,\lambda\in C$ and $\nu\in R.$
The generalization form of the equation (1.2) was presented by Al-Shammery et al.[3] in the following manner

$$D^{a}_{\mu}h(\mu)=\lambda\int_{0}^{\mu}\psi^{\delta}h(\mu-\psi)e^{i\nu\psi}d\psi+be^{i\nu\mu},~{}~{}~{}~{}~{}0\leq\mu\leq 1,$$ (1.10)

where $b,\lambda\in C$ $\nu\in R.$and $\delta>-1$
Al-Shammery et al. [4] established a new form of the equation (1.3) by including function $\Phi(.)$

$$D^{a}_{\mu}h(\mu)=\lambda\int_{0}^{\mu}\psi^{\delta}h(\mu-\psi)\Phi(\beta,\delta+1;i\nu\psi)d\psi+b\Psi(\beta^{\prime},1;i\nu\psi),~{}~{}~{}~{}~{}0\leq\mu\leq 1,$$ (1.11)

where $b,\lambda\in C$ $\nu,\beta,\beta^{\prime}\in R,a>0$ and $\delta>-1$. Now, in this paper we will deal with equation (1.11) with the function introduced by Singh and Porwal [5] known as the incomplete Mittag-Leffler function and whose definition is as follows

$$E_{a,b}^{[\delta,x]}=\sum_{k=0}^{\infty}\frac{[\delta,x]_{k}}{\Gamma(ak+b)}\frac{z^{k}}{k!}$$ (1.12)

$$E_{a,b}^{(\delta,x)}=\sum_{k=0}^{\infty}\frac{(\delta,x)_{k}}{\Gamma(ak+b)}\frac{z^{k}}{k!},$$ (1.13)

where $a,b,\delta\in C$; $\Re(a)>0,\Re(b)>0,\Re(\delta)>0$ and $[\delta,x]_{k}$ and $(\delta,x)_{k}$ represent incomplete Pocchammer symbol and these incomplete Pocchammer symbols satisfy the following decomposition

$$(\lambda;x)_{\nu}+[\lambda;x]_{\nu}=(\lambda)_{\nu}~{}~{}~{}~{}~{}~{}~{}~{}(\lambda,\nu\in C;x\geq 0).$$ (1.14)

Incomplete Wright Function:
Incomplete generalized hypergeometric function [5] is defined as

$${{}_{q}{\overline{\Psi}}_{q}(z)}={{}_{q}{\overline{\Psi}}_{q}}\left[\begin{array}[]{cc}[a_{1},\alpha_{1},x]...(a_{p},\alpha_{p})\\ (b_{1},\beta_{1},x]...(b_{q},\beta_{q})\end{array}|z\right]$$

$$=\sum_{k=0}^{\infty}\frac{\Gamma(a_{1}+\alpha_{1}k,x)\Gamma(a_{2}+\alpha_{2}k)...\Gamma(a_{p}+\alpha_{p}k)}{\Gamma(b_{1}+\beta_{1}k,x)\Gamma(b_{2}+\beta_{2}k)...\Gamma(b_{q}+\beta_{q}k)}\frac{z^{k}}{k!}$$ (1.15)

$${{}_{q}{\underline{\Psi}}_{q}(z)}={{}_{q}{\underline{\Psi}}_{q}}\left[\begin{array}[]{cc}(a_{1},\alpha_{1},x)...(a_{p},\alpha_{p})\\ (b_{1},\beta_{1},x]...(b_{q},\beta_{q})\end{array}|z\right]$$

$$\sum_{k=0}^{\infty}\frac{\gamma(a_{1}+\alpha_{1}k,x)\Gamma(a_{2}+\alpha_{2}k)...\Gamma(a_{p}+\alpha_{p}k)}{\gamma(b_{1}+\beta_{1}k,x)\Gamma(b_{2}+\beta_{2}k)...\Gamma(b_{q}+\beta_{q}k)}\frac{z^{k}}{k!}$$ (1.16)

With decomposition formula (1.14) incomplete Wright function (1.12) and (1.13) as classical Wright function ${}_{p}\Psi_{q}(z)$ [1] and this particular function is an entire function if there hold the condition

$$\sum_{j=1}^{q}\beta_{j}-\sum_{i=1}^{p}\alpha_{j}>1.$$ (1.17)

2 Theorems and Results

Theorem 2.1

Let $b,c,\omega,\in C,\rho\in\Re,a>0,\rho>0,\Re(b)>0$ and $g(t)$ is assumed to be continuous on every finite interval $\left[0,T\right]$,$0<T<\infty$, and of the exponential order $e^{\rho\mu}$, when $\mu\to\infty$. Then for the generalized Free Electron Laser equation for the fractional integro-differential equation of Volterra type

$$D^{a}_{\mu}h(\mu)=\omega\int_{0}^{\mu}t^{b-1}E^{[c;x]}_{\rho,b}(i\zeta t^{\rho})h(\mu-t)dt+\delta g(t)~{}~{}~{}~{}(0\leq\mu\leq 1).$$ (2.1)

together with the initial conditions

$$D^{a-r}_{\mu}h(\mu)|_{\mu}=0=b_{r},r=1,...,n=-[-\Re(a)],\left(n-1<a\leq n\right);n\in N,$$ (2.2)

where $b_{1},b_{2},b_{3},....,b_{r}\in\Re$, there exist a unique continuous solution given by

$$h(\mu)=\sum_{r=1}^{n}b_{r}y_{r}(\mu)+\delta\int_{0}^{\mu}\aleph(\mu-t)g(t)dt$$ (2.3)

where

$$y_{r}(\mu)=\sum_{k=0}^{\infty}\omega^{k}\mu^{a+(a+b)k-r}E^{[c;x]k}_{\rho,1+a+(a+b)k-r}(i\zeta\mu^{\rho})\\$$ (2.4)

$$y_{r}(\mu)=\sum_{k=0}^{\infty}\omega^{k}\mu^{a+(a+b)k-r}\frac{1}{\Gamma(ck)}{{}_{1}}\overline{\Psi}_{1}\left[\begin{array}[]{cc}[ck,1;xk]\\ (1+a+(a+b)k-r,\rho)\end{array}|i\zeta\mu^{\rho}\right]$$ (2.5)

$$\aleph(x)=\sum_{k=0}^{\infty}\omega^{k}x^{a+(a+b)k-1}E^{[c;x]k}_{\rho,a+(a+b)k}(i\zeta x^{\rho})$$ (2.6)

$$\aleph(x)=\sum_{k=0}^{\infty}\omega^{k}x^{a+(a+b)k-1}\frac{1}{\Gamma(ck)}{{}_{1}}\overline{\Psi}_{1}\left[\begin{array}[]{cc}[ck,1;xk]\\ (a+(a+b)k,\rho)\end{array}|i\zeta\mu^{\rho}\right]$$ (2.7)

Proof.

Taking Laplace on both side of the integro-differential equation (2.1) and using (1.12) and (1.5), we find that

$$L\{D^{a}_{\mu}h{\left(\mu\right)}\}=L\{\omega\int^{\mu}_{0}t^{b-1}h(\mu-t)E^{[c;x]}_{\rho,b}(i\zeta t^{\rho})dt+\delta g(t)\}$$ (2.8)

$$s^{a}H(s)-\sum_{r=0}^{n}s^{r-1}D^{a-r}_{\mu}h(\mu)|_{\mu=0}=\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}H(s)+\delta G(s)$$ (2.9)

$$\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]H(s)=\sum_{r=0}^{n}s^{r-1}D^{a-r}_{\mu}h(\mu)|_{\mu=0}+\delta G(s)$$ (2.10)

$$\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]H(s)=\sum_{r=0}^{n}s^{r-1}b_{r}+\delta G(s),$$ (2.11)

where $D^{a-r}_{\mu}h(\mu)|_{\mu=0}$.

$$H(s)=\sum_{r=1}^{n}s^{r-1}b_{r}\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}+\delta G(s)\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$H(s)=\sum_{r=1}^{n}s^{r-1}b_{r}s^{-a}\left[1-\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$+\delta G(s)s^{-a}\left[1-\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$=\sum_{r=1}^{n}s^{r-1}b_{r}s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}$$

$$+\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}$$ (2.12)

where $|\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}|<1$. Taking the inverse Laplace transform on both sides of (2.12), we get

$$L^{-1}\{H(s)\}=L^{-1}\{\sum_{r=1}^{n}s^{r-1}b_{r}s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}\}$$

$$+L^{-1}\{\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}\}$$

$$h(s)=L^{-1}\{\sum_{r=1}^{n}s^{r-1}b_{r}s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}\sum_{n=0}^{\infty}(i\zeta s^{-\rho})^{n}([c;x]k)_{n}\frac{1}{n!}\}$$

$$+L^{-1}\{\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}\sum_{n=0}^{\infty}(i\zeta s^{-\rho})^{n}([c;x]k)_{n}\frac{1}{n!}\}$$

$$h(s)=\sum_{r=1}^{n}b_{r}\sum_{k=0}^{\infty}\omega^{k}\sum_{n=0}^{\infty}(i\zeta)^{n}([c;x]k)_{n}\frac{1}{n!}L^{-1}\{s^{r-a-(a+b)k-\rho n}\}$$

$$+\delta\sum_{k=0}^{\infty}\omega^{k}\sum_{n=0}^{\infty}(i\zeta)^{n}([c;x]k)_{n}\frac{1}{n!}L^{-1}\{G(s)s^{-a-(a+b)k-\rho n}\}$$ (2.13)

Using $L^{-1}\{\frac{1}{s^{n}}\}=\frac{t^{n-1}}{(n-1)!}$ in the above expression, then after some calculation, we arrive at

$$=\sum_{r=1}^{n}b_{r}\sum_{k=0}^{\infty}\omega^{k}\mu^{a+(a+b)k-r}E^{[c;x]k}_{\rho,1+a+(a+b)k-r}(i\zeta\mu^{\rho})$$

$$+\delta\int_{0}^{\mu}\sum_{k=0}^{\infty}\omega^{k}(\mu-t)^{a+(a+b)k-1}E^{[c;x]k}_{\rho,a+(a+b)k}(i\zeta(\mu-t)^{\rho})g(t)dt.$$ (2.14)

Which is the required result and by using (1.15) in (2.14), we can achieve (2.5) and (2.7).
∎

Theorem 2.2

Let $b,c,\omega,\in C,\rho\in\Re,a>0,\rho>0,\Re(b)>0$ and $g(t)$ is assumed to be continuous on every finite interval $\left[0,T\right]$,$0<T<\infty$, and of the exponential order $e^{\rho\mu}$, when $\mu\to\infty$. Then for the generalized Free Electron Laser equation for the fractional integro differential equation of Voltera type

$$\frac{d^{a}}{d\mu^{a}}h(\mu)=\omega\int_{0}^{\mu}t^{b-1}E^{[c;x]}_{\rho,b}(i\zeta t^{\rho})h(\mu-t)dt+\delta g(t)~{}~{}~{}~{}(0\leq\mu\leq 1).$$ (2.15)

together with the initial conditions

$$\frac{d^{a}}{d\mu^{a}}h(\mu)|_{\mu}=0=a_{r},r=1,...,n-1;\left(n-1<a\leq n\right);n\in N,$$ (2.16)

where $a_{1},a_{2},a_{3},....,a_{n-1}\in\Re$, there exist a unique continuous solution given by

$$h(\mu)=\sum_{r=0}^{n-1}a_{r}y_{r}(\mu)+\delta\int_{0}^{\mu}\aleph(\mu-t)g(t)dt$$ (2.17)

where

$$y_{r}(\mu)=\sum_{k=0}^{\infty}\omega^{k}\mu^{r+(a+b)k}E^{[c;x]k}_{\rho,1+(a+b)k+r}(i\zeta\mu^{\rho})$$ (2.18)

$$y_{r}(\mu)=\sum_{k=0}^{\infty}\omega^{k}\mu^{(a+b)k+r}\frac{1}{\Gamma(ck)}{{}_{1}}\overline{\Psi}_{1}\left[\begin{array}[]{cc}[ck,1;xk]\\ (1+(a+b)k+r,\rho)\end{array}|i\zeta\mu^{\rho}\right]$$ (2.19)

$$\aleph(x)=\sum_{k=0}^{\infty}\omega^{k}x^{a+(a+b)k-1}E^{[c;x]k}_{\rho,a+(a+b)k}(i\zeta x^{\rho})$$ (2.20)

$$\aleph(x)=\sum_{k=0}^{\infty}\omega^{k}x^{a+(a+b)k-1}\frac{1}{\Gamma(ck)}{{}_{1}}\overline{\Psi}_{1}\left[\begin{array}[]{cc}[ck,1;xk]\\ (a+(a+b)k,\rho)\end{array}|i\zeta\mu^{\rho}\right]$$ (2.21)

Proof.

Taking Laplace on both side of the integro-differential equation (2.15) and using (1.12) and (1.5), we find that

$$L\{\frac{d^{a}}{d\mu^{a}}h{\left(\mu\right)}\}=L\{\omega\int^{\mu}_{0}t^{b-1}h(\mu-t)E^{[c;x]}_{\rho,b}(i\zeta t^{\rho})dt+\delta g(t)\}.$$

$$s^{a}H(s)-\sum_{r=0}^{n-1}s^{a-r-1}\frac{d^{a}}{d\mu^{a}}h(\mu)|_{\mu=0}=\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}H(s)+\delta G(s).$$

$$\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]H(s)=\sum_{r=0}^{n-1}s^{a-r-1}\frac{d^{a}}{d\mu^{a}}h(\mu)|_{\mu=0}+\delta G(s).$$

$$\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]H(s)=\sum_{r=0}^{n-1}s^{a-r-1}b_{r}+\delta G(s),$$

where $\frac{d^{a}}{d\mu^{a}}h(\mu)|_{\mu=0}$.

$$H(s)=\sum_{r=0}^{n-1}s^{a-r-1}b_{r}\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}+\delta G(s)\left[s^{a}-\omega s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$H(s)=\sum_{r=0}^{n-1}s^{-r-1}b_{r}\left[1-\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$+\delta G(s)s^{-a}\left[1-\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}\right]^{-1}$$

$$=\sum_{r=0}^{n-1}s^{-r-1}b_{r}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}$$

$$+\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}$$ (2.22)

where $|\omega s^{-a}s^{-b}(1-i\zeta s^{-\rho})^{-[c;x]}|<1$. Taking the inverse Laplace transform on both sides of (2.22), we get

$$L^{-1}\{H(s)\}=L^{-1}\{\sum_{r=0}^{n-1}s^{-r-1}b_{r}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}\}$$

$$+L^{-1}\{\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}(1-i\zeta s^{-\rho})^{-[c;x]k}\}$$

$$h(s)=L^{-1}\{\sum_{r=0}^{n-1}s^{-r-1}b_{r}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}\sum_{n=0}^{\infty}(i\zeta s^{-\rho})^{n}([c;x]k)_{n}\frac{1}{n!}\}$$

$$+L^{-1}\{\delta G(s)s^{-a}\sum_{k=0}^{\infty}\left(\omega s^{-a}s^{-b}\right)^{k}\sum_{n=0}^{\infty}(i\zeta s^{-\rho})^{n}([c;x]k)_{n}\frac{1}{n!}\}$$

$$h(s)=\sum_{r=0}^{n-1}b_{r}\sum_{k=0}^{\infty}\omega^{k}\sum_{n=0}^{\infty}(i\zeta)^{n}([c;x]k)_{n}\frac{1}{n!}L^{-1}\{s^{-r-(a+b)k-\rho n-1}\}$$

$$+\delta\sum_{k=0}^{\infty}\omega^{k}\sum_{n=0}^{\infty}(i\zeta)^{n}([c;x]k)_{n}\frac{1}{n!}L^{-1}\{G(s)s^{-a-(a+b)k-\rho n}\}$$ (2.23)

Using $L^{-1}\{\frac{1}{s^{n}}\}=\frac{t^{n-1}}{(n-1)!}$ in the above expression, then after some calculation, we arrive at

$$=\sum_{r=0}^{n-1}b_{r}\sum_{k=0}^{\infty}\omega^{k}\mu^{r+(a+b)k}E^{[c;x]k}_{\rho,1+(a+b)k+r}(i\zeta\mu^{\rho})$$

$$+\delta\int_{0}^{\mu}\sum_{k=0}^{\infty}\omega^{k}(\mu-t)^{a+(a+b)k-1}E^{[c;x]k}_{\rho,a+(a+b)k}(i\zeta(\mu-t)^{\rho})g(t)dt.$$ (2.24)

By using (1.15) in (2.24), we can achieve (2.19) and (2.21). ∎

Funding. There is no source of funding for this article.
Ethical approval. This article does not contain any studies with human participants or animals.
Availability of data and materials Not applicable.
Conflict of interest. The author declares that this paper involves no conflict of interest.