nlin0606070/PSI.tex
1: \documentclass[12pt]{article}
2: 
3: \voffset=-11mm
4: \hoffset=-11 mm
5: \textheight=210mm \textwidth=160mm
6: 
7: %%%\def\bfkappa{\mathord{\mbox{\boldmath $\kappa$}}}
8: %\def\bfkappa{{\vec{\kappa}}}
9: %\def\bfkappa{{\bf \kappa}}
10: \def\bfkappa{\mathop{\mbox{\boldmath $\kappa$}}}
11: \def\bfeta{\mathord{\mbox{\boldmath $\eta$}}}
12: \def\bfnabla{\mathord{\mbox{\boldmath $\nabla$}}}
13: 
14: \renewcommand{\baselinestretch}{1.2}
15: 
16: \newcommand{\be}{\begin{equation}}
17: \newcommand{\ee}{\end{equation}}
18: \newcommand{\bi}[1]{\vspace{-3mm} \bibitem{#1}}
19: 
20: %%\usepackage{epsfig}
21: \usepackage{epsfig,amsmath}
22: \usepackage{graphics}
23: 
24: 
25: \begin{document}
26: 
27: %%%\today \\
28: \begin{center}
29: Journal of Physics A. Vol.39. No.26. (2006) pp.8395-8407.
30: \end{center}
31: 
32: \begin{center}
33: {\Large \bf Psi-Series Solution of Fractional Ginzburg-Landau Equation}
34: \vskip 5 mm
35: 
36: {\large \bf Vasily E. Tarasov} \\
37: 
38: \vskip 3mm
39: 
40: {\it Skobeltsyn Institute of Nuclear Physics, \\
41: Moscow State University, Moscow 119992, Russia } \\
42: {E-mail: tarasov@theory.sinp.msu.ru}
43: \end{center}
44: 
45: \vskip 11 mm
46: 
47: 
48: \begin{abstract}
49: One-dimensional Ginzburg-Landau equations
50: with derivatives of noninteger order are considered.
51: Using psi-series with fractional powers, the solution of 
52: the fractional Ginzburg-Landau (FGL) equation is derived.
53: The leading-order behaviours of solutions about an arbitrary 
54: singularity, as well as their resonance structures, 
55: have been obtained.
56: It was proved that fractional equations of order $\alpha$ 
57: with polynomial nonlinearity of order $s$  
58: have the noninteger power-like behavior of order $\alpha/(1-s)$ 
59: near the singularity.
60: \end{abstract}
61: 
62: PACS numbers: 05.45.-a; 45.10.Hj \\
63: 
64: %%% 05.45.-a Nonlinear dynamics and nonlinear dynamical systems 
65: %%%45.10.Hj  Perturbation and fractional calculus methods
66: 
67: 
68: \section{Introduction}
69: 
70: Differential equations that contain derivatives 
71: of noninteger order \cite{SKM,OS} are called fractional equations 
72: \cite{Podlubny,KST}. 
73: The interest to fractional equations has been growing continually 
74: during the last few years because of numerous applications. 
75: In a fairly short period of time the areas of 
76: applications of fractional calculus have become broad.
77: For example, the derivatives and integrals of fractional order 
78: are used in chaotic dynamics \cite{Zaslavsky1,Zaslavsky2},
79: material sciences \cite{Hilfer,C2,Nig1}, 
80: mechanics of fractal and complex media \cite{Mainardi,Media},
81: quantum mechanics \cite{Laskin,Naber}, 
82: physical kinetics \cite{Zaslavsky1,Zaslavsky7,SZ,ZE},
83: plasma physics \cite{CLZ,Plasma2005}, 
84: astrophysics \cite{CMDA},
85: %%and many others physical topics related to 
86: long-range dissipation \cite{GM}, 
87: non-Hamiltonian mechanics \cite{nonHam,FracHam},
88: long-range interaction \cite{LZ,TZ3,KZT}, 
89: anomalous diffusion and transport theory 
90: \cite{Zaslavsky1,Montr,Uch}.
91: 
92: The fractional generalization of the Ginzburg-Landau equation
93: was suggested in \cite{Zaslavsky6}.
94: This equation can be used to describe the dynamical processes 
95: in continuums with fractal dispersion and 
96: the media with fractal mass dimension \cite{Physica2005,Mil,TZ2}. 
97: In this paper, we generalize the psi-series 
98: approach to the fractional differential equations.
99: As an example, we consider a solution of 
100: the fractional Ginzburg-Landau (FGL) equation. 
101: We derive the psi-series for the one-dimensional FGL equation.
102: The leading-order behaviours of solutions about an arbitrary 
103: singularity, as well as their resonance structures, 
104: have been obtained.
105: 
106: In section 2, we recall the appearance of the Ginzburg-Landau equation 
107: with fractional derivatives.
108: In section 3, the singular behavior of the FGL equation is considered. 
109: In section 4, we discuss the powers of series terms 
110: that have arbitrary coefficients that are called 
111: the resonances or Kovalevskaya exponents.
112: In section 5, we derive the psi-series and recurrence relations
113: for one-dimensional FGL equation with
114: rational order $\alpha$ ($1<\alpha <2$).
115: In section 6, the example of an FGL equation with order $\alpha=3/2$
116: is suggested. In section 7, the next to singular behaviour 
117: for arbitrary (rational or irrational) order is discussed.
118: Finally, a short conclusion is given in section 8.
119: 
120: 
121: 
122: 
123: 
124: 
125: \section{Fractional Ginzburg-Landau (FGL) equation}
126:  
127: 
128: Let us recall the appearance of the nonlinear parabolic equation 
129: \cite{Leon,Light,Kad,ZS}, and 
130: the FGL equation \cite{Zaslavsky6,Physica2005,TZ2}.
131: Consider wave propagation in some media and 
132: present the wave vector $\bf k$ in the form
133: \begin{equation}
134: {\bf k} = {\bf k}_0 + {\bfkappa} = {\bf k}_0 + {\bfkappa}_{\parallel}
135: + {\bfkappa}_{\perp}, 
136: \label{eq:36}
137: \end{equation}
138: where ${\bf k}_0$ is the unperturbed wave vector and subscripts
139: $(\parallel ,\perp )$ are taken respectively to the direction of ${\bf k}_0$. 
140: A symmetric dispersion law $\omega = \omega (k)$ for
141: $\kappa \ll k_0$ can be written as
142: \be \label{PE1}
143: \omega (k)= \omega (|{\bf k}|) \approx 
144: \omega (k_0)+ \ v_g \ (|{\bf k}| - k_0 )+
145: {1 \over 2} v^{\prime }_g \ (|{\bf k}| - k_0 )^2 ,
146: \ee
147: where
148: \be \label{PE2}
149: v_g = \left(\frac{\partial\omega}{\partial k} \right)_{k=k_0} , \quad
150: v^{\prime}_g=\left(\frac{\partial^2\omega}{\partial k^2}\right)_{k=k_0} ,
151: \ee
152: and
153: \be \label{PE3}
154: |{\bf k}|=|{\bf k}_0 + {\bfkappa}|=
155: \sqrt{({\bf k}_0+\kappa_{\parallel})^2+\kappa^2_{\perp}} \approx
156: k_0+\kappa_{\parallel}+\frac{1}{2k_0}\kappa^2_{\perp}.
157: \ee
158: Substitution of (\ref{PE3}) into (\ref{PE1}) gives
159: \be
160: \omega (k) \approx \omega_0 + v_g\kappa_{\parallel} +
161: {v_g\over 2k_0} \kappa_{\perp}^2 +
162: \frac{v^{\prime}_g}{2} \kappa_{\parallel}^2 ,
163: \label{eq:37}
164: \ee
165: where $\omega_0=\omega(k_0)$. 
166: Expression (\ref{eq:37}) in the dual space ("momentum representation") 
167: corresponds to the following equation in the coordinate space:
168: \begin{equation}
169: i {\partial Z \over \partial t} =\omega_0 Z- 
170: i v_g {\partial Z \over \partial x}
171:  - {v_g \over 2k_0 } \Delta_{\perp} Z-
172: {v^{\prime}_g \over 2 } \Delta_{\parallel} Z
173:  \label{eq:38}
174: \end{equation}
175: with respect to the field $Z=Z(t,x,y,z)$, 
176: where $x$ is along ${\bf k}_0$, 
177: and we use the operator correspondence between the dual space 
178: and usual space-time:
179: \[
180: \omega (k)  \ \longleftrightarrow \
181:  i {\partial\over\partial t} , \quad
182: \kappa_{\parallel} \ \longleftrightarrow \ 
183:  -  i {\partial\over\partial x} , \] 
184: \be
185: {(\bfkappa}_{\perp})^2 \ \longleftrightarrow \ - \Delta_{\perp}=
186: - {\partial^2 \over\partial y^2 }-{\partial^2 \over\partial z^2 } , \quad
187: {(\bfkappa}_{\parallel})^2 \ \longleftrightarrow \ 
188: - \Delta_{\parallel}=- {\partial^2 \over\partial x^2 } .
189: \label{eq:39}
190: \ee
191: A generalization to the nonlinear case can be carried out similarly to
192: (\ref{eq:37}) through a nonlinear dispersion law dependence on the wave
193: amplitude:
194: \begin{equation} 
195: \omega = \omega (k,|Z|^2 ) \approx\omega (k,0) + b|Z|^2
196:  = \omega (|{\bf k}|) + b|Z|^2
197:  \label{eq:40}
198: \end{equation}
199: with some constant 
200: $b = \partial\omega (k, |Z|^2 )/\partial |Z|^2$ at $|Z| = 0$. 
201: In analogy to (\ref{eq:38}), we obtain from (\ref{eq:37}) and (\ref{eq:39}): 
202: \begin{equation}
203: i {\partial Z \over \partial t} =\omega(k_0) Z - 
204: i v_g {\partial Z \over \partial x}
205:  - {v_g \over 2k_0 } \Delta_{\perp} Z-
206: {v^{\prime}_g \over 2 } \Delta_{\parallel} Z+ b|Z|^2 Z .
207:  \label{eq:41}
208: \end{equation}
209: This equation is known as the nonlinear parabolic equation 
210: \cite{Leon,Light,Kad,ZS}. 
211: The change of variables from $(t,x,y,z)$ to $(t,x-v_gt,y,z)$ gives
212: \begin{equation}
213: -i {\partial Z \over \partial t}
214:  = {v_g \over 2k_0} \Delta_{\perp} Z +
215: {v^{\prime}_g \over 2} \Delta_{\parallel} Z 
216: - \omega(k_0) Z - b|Z|^2 Z 
217: %%%\quad \Delta_{\parallel}=\frac{\partial^2}{\partial \xi^2}
218:  \label{eq:42}
219: \end{equation}
220: that is also known as the nonlinear Schr\"{o}dinger (NLS) equation.
221: 
222: Wave propagation in media with fractal properties can be easily
223: generalized by rewriting the dispersion law (\ref{eq:37}), (\ref{eq:40})
224: in the following way \cite{Zaslavsky6}:
225: \begin{equation}
226:  \omega (k,|Z|^2 )= \omega (k_0 ,0) + v_g \kappa_{\parallel} +
227: g_1 ({\bfkappa}_{\perp}^2 )^{\alpha /2} +
228: g_2 ({\bfkappa}_{\parallel}^2 )^{\beta /2}+ b|Z|^2,
229: \quad (1<\alpha, \beta <2)
230:  \label{eq:43}
231: \end{equation}
232: with new constants $g_1$, $g_2$.
233: 
234: Using the connection between Riesz fractional derivative 
235: and its Fourier transform \cite{SKM} 
236: \begin{equation}
237: (-\Delta_{\perp} )^{\alpha /2} 
238: \longleftrightarrow
239: ({\bfkappa}_{\perp}^2 )^{\alpha /2} ,
240: \quad
241: (-\Delta_{\parallel} )^{\beta /2} 
242: \longleftrightarrow  
243: ({\bfkappa}_{\parallel}^2 )^{\beta /2} ,
244:  \label{eq:44}
245: \end{equation}
246: we obtain from (\ref{eq:43})
247: \begin{equation}
248: i {\partial Z \over \partial t} =- iv_g {\partial Z \over \partial x}
249: + g_1 (-\Delta_{\perp} )^{\alpha /2}  Z 
250: + g_2 (-\Delta_{\parallel} )^{\beta /2}  Z 
251: + \omega_0 Z + b|Z|^2 Z,
252:   \label{eq:45}
253: \end{equation}
254: where $Z=Z(t,x,y,z)$. 
255: By changing the variables from $(t,x,y,z)$ to $(t,\xi,y,z)$, 
256: $\xi=x-v_gt$, and using 
257: \be
258: (-\Delta_{\parallel} )^{\beta /2}=
259: \frac{\partial^{\beta} }{\partial |x|^{\beta}}=
260: \frac{\partial^{\beta} }{\partial |\xi|^{\beta}}, 
261: \ee
262: we obtain  from (\ref{eq:45}) equation
263: \begin{equation}
264: i {\partial Z \over \partial t} =
265:  g_1 (-\Delta_{\perp} )^{\alpha /2}  Z 
266: + g_2 (-\Delta_{\parallel} )^{\beta /2}  Z 
267: + \omega_0 Z + b|Z|^2 Z,
268: \label{eq:46}
269: \end{equation}
270: that can be called the fractional nonlinear parabolic equation.
271: For $g_2=0$ we get the
272: nonstationary FGL equation (fractional NLS equation)
273: suggested in \cite{Zaslavsky6}.
274: Let us comment on the physical structure of (\ref{eq:46}).
275: The first term on the right-hand side is related to 
276: wave propagation in media with fractal properties. 
277: The fractional derivative can also appear
278: as a result of long-range interaction \cite{LZ,TZ3,KZT}. 
279: Other terms on the right-hand-side of equations (\ref{eq:45}) 
280: and (\ref{eq:46}) correspond to  wave interaction due to
281: the nonlinear properties of the media. Thus, equation (\ref{eq:46}) 
282: can describe fractal processes of self-focusing and related issues.
283: 
284: We may consider one-dimensional simplifications of (\ref{eq:46}), i.e., 
285: \begin{equation}
286: i {\partial Z \over \partial t}=
287: g_2 \frac{\partial^{\beta} Z}{\partial |\xi|^{\beta}}
288: + \omega_0 Z + b|Z|^2 Z,
289: \label{eq:47a}
290: \end{equation}
291: where $Z=Z(t,\xi)$, $\xi=x-v_gt$, or the equation
292: \begin{equation}
293: i {\partial Z \over \partial t} =
294: g_1 \frac{\partial^{\alpha} Z}{\partial |z|^{\alpha}}
295: + \omega_0 Z + b|Z|^2 Z,
296: \label{eq:47b}
297: \end{equation}
298: where $Z=Z(t,z)$.
299: We can reduce (\ref{eq:47b}) to the case of a propagating wave 
300: \be
301: Z=Z(z-v_gt)\equiv Z(\eta) .
302: \ee
303: For the real field $Z$, Eq. (\ref{eq:47b}) becomes
304: \be \label{eq:49}
305: g_1\frac{d^{\alpha}Z}{d |\eta|^{\alpha}}+c\frac{dZ}{d\eta}+\omega_0 Z+bZ^3=0,
306: \quad \eta=z-v_g t ,
307: \ee
308: where $c=iv_g$.
309: This equation takes the form of the fractional generalization of
310: the Ginzburg-Landau equation, when $v_g=0$.
311: 
312: It is well known that the nonlinear term in equations of type
313: (\ref{eq:42})  leads to a steepening of the solution and its
314: singularity. The steepening process may be stopped by a diffusional or
315: dispersional term, i.e. by a higher derivative term. A similar phenomenon
316: may appear for the fractional nonlinear equations (\ref{eq:49}). 
317: 
318: 
319: 
320: 
321: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
322: \section{Singular behavior of FGL equation}
323: 
324: 
325: There is an approach to the question of integrability which 
326: is not concerned with the display of explicit functions, 
327: but with the demonstration of a specific property. 
328: This is the existence of Laurent series for each of the dependent variables. 
329: The series may not be summable to an explicit form, but does represent 
330: an analytic function. The essential feature of this 
331: Laurent series is that it is an expansion about a particular type of
332: movable singularity, i.e., a pole. 
333: Consequently the existence of these Laurent series is intimately 
334: concerned with the singularity analysis of differential 
335: equations initiated about a century ago by Painleve, Gambier and Garnier 
336: \cite{Ince} and continued since by many workers
337: including Bureau \cite{3} and Cosgrove et al \cite{4}.
338: 
339: The connection of this type of singular behavior and 
340: the solution of partial differential equations by the method of the 
341: inverse scattering transform was noticed by Ablowitz et al \cite{Ablowitz}
342: who developed an algorithm, called the ARS algorithm, to test whether 
343: the solution of an ordinary differential equation was expressible 
344: in terms of a Laurent expansion. 
345: If this was the case, the ordinary differential equation 
346: was said to pass the Painleve test and was conjectured to be integrable. 
347: Under more precise conditions Conte \cite{8} has shown that 
348: the equation is integrable. Psi-series solutions of differential 
349: equations are considered in \cite{Tabor1,Tabor2,BSV,CTW}.
350: 
351: In the paper, we consider the fractional equation 
352: \be \label{FGL}
353: gD^{\alpha}_x Z(x)+cD^{1}_x Z(x)+aZ(x) +bZ^3(x)=0,
354: \ee
355: where $1<\alpha < 2$, and $D^{\alpha}_x$ is 
356: the fractional Riemann-Liouville derivative:
357: \be
358: D^{\alpha}_x Z(x)=\frac{1}{\Gamma (m-\alpha)} \frac{d^m}{d x^m}  
359: \int^x_{x_0}  dy \frac{Z(y)}{(x-y)^{\alpha-m+1}} . \ee
360: Here, $m$ is the first whole number greater than or equal to $\alpha$.
361: In our case $m=2$. 
362: We detect possible singular behavior in the solution of a 
363: differential equation by means of the leading-order analysis. 
364: 
365: To determine the leading-order behavior, we set
366: \be \label{Zfp}
367: Z(x)=f(x-x_0)^{p} ,
368: \ee
369: where $x_0$ is an arbitrary constant (the location of the singularity).  
370: %%%is the location of the supposed singularity, movable singularity,
371: Then, we substitute (\ref{Zfp}) into the fractional differential 
372: equation (\ref{FGL}) and look for two or more dominant terms.
373: The detection of which terms are dominant is identical to the determination
374: of which terms in an equation are self-similar.
375: 
376: Substituting (\ref{Zfp}) into equation (\ref{FGL}), and
377: using the relation
378: \be \label{Da}
379: D^{\alpha}_x x^p=
380: \frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} x^{p-\alpha} , \quad (p>-1) ,
381: \ee
382: we get
383: \be
384: gf \frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} (x-x_0)^{p-\alpha}+
385: cpf(x-x_0)^{p-1}+af (x-x_0)^p+ bf^3 (x-x_0)^{3p}=0.
386: \ee
387: If $1< \alpha < 2$, then $p-\alpha < p-1$. For the dominant terms, 
388: \be \label{dominant}
389: gf \frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} (x-x_0)^{p-\alpha}+
390: bf^3 (x-x_0)^{3p}=0.
391: \ee
392: As a result, we obtain
393: \be \label{p}
394: p-\alpha=3p,
395: \ee
396: \be \label{f}
397: g\frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} + bf^2 =0.
398: \ee
399: Equation (\ref{p}) gives
400: \be 
401: p=-\frac{\alpha}{2}.
402: \ee
403: If $1<\alpha<2$, then $-1<p<-1/2$. 
404: Therefore the leading-order singular behavior is found:
405: \be
406: Z(x)=f(x-x_0)^{-\alpha/2}, \quad 
407: f^2+\frac{g\Gamma(1-\alpha/2)}{b\Gamma(1-3\alpha/2)}=0,
408: \ee
409: and the singularity is a pole of order $\alpha/2$.
410: Evidently our psi-series starts at
411: $(x-x_0)^{-\alpha/2}$.
412: The resonance conditions and psi-series 
413: is considered in the next sections.
414: 
415: As a result, we get that
416: fractional differential equations of order $\alpha$ 
417: with polynomial nonlinearity of order $s$
418: have the noninteger power behavior of order $\alpha/(1-s)$ 
419: near the singularity.
420: 
421: 
422: \section{Resonance terms of FGL equation}
423: 
424: 
425: The powers of $(x-x_0)$ that have arbitrary coefficients
426: are called the resonances or Kovalevskaya's exponents. 
427: To find resonance, we consider the substitution
428: \be \label{Zfplr}
429: Z(x)=f (x-x_0)^p +l (x-x_0)^{p+r}, 
430: \ee
431: and find the values $r$. 
432: In equation (\ref{Zfplr}) parameters $p$ and $f$ are defined  by
433: \be \label{p-f0}
434: p=-\frac{\alpha}{2}, \quad
435: f=\sqrt{-\frac{g\Gamma(1-\alpha/2)}{b\Gamma(1-3\alpha/2)}} .
436: \ee
437: Substitution of Eq. (\ref{Zfplr}) into (\ref{FGL}) gives
438: \[ gf \frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} (x-x_0)^{p-\alpha}+
439: cpf(x-x_0)^{p-1}+af (x-x_0)^p+ bf^3 (x-x_0)^{3p}+\]
440: \[ +g l \frac{\Gamma(p+r+1)}{\Gamma(p+r+1-\alpha)} (x-x_0)^{p+r-\alpha}+
441: cpl(x-x_0)^{p+r-1}+al (x-x_0)^{p+r}+ bl^3 (x-x_0)^{3p+3r}+ \]
442: \be \label{29}
443: +3bl^2 f (x-x_0)^{2p+3r}+3bl f^2 (x-x_0)^{p+3r}=0 .
444: \ee
445: Using equation (\ref{p-f0}), and considering 
446: the linear with respect to $l$ terms of (\ref{29}), we have
447: \be \label{res}
448: \frac{\Gamma(1+r-\alpha/2)}{\Gamma(1+r-3\alpha/2)}-
449: 3 \frac{\Gamma(1-\alpha/2)}{\Gamma(1-3\alpha/2)}=0 .
450: \ee
451: This relation allows us to derive the values of $r$.
452: Equation (\ref{res}) can be directly derived by using the recurrence relations.
453: In the general case, the values of $r$ can be irrational or complex numbers.
454: The solution of the FGL equation with $1<\alpha \le 2$ have 
455: two arbitrary parameters.
456: Therefore, we must have two values of $r$ that are the 
457: solutions of equation (\ref{res}). 
458: It is interesting to note that (\ref{res}) gives
459: two real values of $r$ only for 
460: \be
461: \alpha >\alpha_0 ,
462: \ee
463: where
464: \be
465: \alpha_0 \approx 1.3005888986 .
466: \ee
467: The order $\alpha_0$ is an universal value that does not depends on
468: values of parameters $g$, $a$, $b$, $c$ of the FGL equation (\ref{FGL}).
469: 
470: The plots of the function
471: \be
472: y(r)=\frac{\Gamma(1+r-\alpha/2)}{\Gamma(1+r-3\alpha/2)}-
473: 3 \frac{\Gamma(1-\alpha/2)}{\Gamma(1-3\alpha/2)} 
474: \ee
475: are shown in figures 1 and 2.
476: The solutions of equation (\ref{res}) correspond to the points of 
477: intersections with the horizontal axis. 
478: 
479: As a result, the nature of the resonances is summarized as follows: \\
480: 1) For $\alpha$ such that $1< \alpha < \alpha_0$ the values of $r$
481: are complex or $r <-\alpha/2$. \\
482: 2) For $\alpha$ such that $\alpha_0 < \alpha <2$, we have two the real 
483: values of $r$. Note that for $\alpha_0 <\alpha <1.9999999995$,
484: the values $r$  satisfy the inequality $r<6.426$.
485: 
486: 
487: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
488: %%%   Figures
489: 
490: \begin{figure}[tbh]
491: \begin{center}
492: \resizebox{9cm}{!}{\includegraphics{res1.eps}}
493: \caption{\it Plot for the order $\alpha=1.30$. }  
494: \label{res1}
495: \end{center}
496: \end{figure}
497: \begin{figure}[tbh]
498: \begin{center}
499: \resizebox{9cm}{!}{\includegraphics{res2.eps}}
500: \caption{\it Plot for the order $\alpha=1.31$. }  
501: \label{res2}
502: \end{center}
503: \end{figure}
504: 
505: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
506: 
507: 
508: 
509: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
510: \section{Psi-Series for FGL Equation of Rational Order}
511: 
512: Let us consider the psi-series and recurrence relations
513: for the one-dimensional FGL equation (\ref{FGL}), 
514: where the order $\alpha$ is a rational number such that:
515: \be
516: \alpha=\frac{m}{n}, \quad (1<\alpha <2).
517: \ee
518: 
519: Following a standard procedure \cite{Tabor1}, we substitute
520: \be \label{Zsum1}
521: Z(x)=\frac{1}{(x-x_0)^{\alpha/2}}
522: \sum^{\infty}_{k=0} e_k \phi^k(x-x_0) ,
523: \ee
524: where 
525: \be
526: e_0=f=\sqrt{-\frac{g\Gamma(1-\alpha/2)}{b\Gamma(1-3\alpha/2)}}, 
527: \ee
528: into the fractional Ginzburg-Landau equation (\ref{FGL}).
529: Note that the coefficient $e_0$ is a real number for two cases: 
530: (1) $g/b\ge 0$ and $1<\alpha<4/3$; \ (2) $g/b\le0$ and $4/3 <\alpha <2$.
531: 
532: For the rational order $\alpha=m/n$, 
533: we suggest to use the $\phi$-function in the form
534: \be 
535: \phi(x-x_0)=(x-x_0)^{1/2n} .
536: \ee
537: Then 
538: \be \label{Zsum2}
539: Z(x)=\frac{1}{(x-x_0)^{\alpha/2}}
540: \sum^{\infty}_{k=0} e_k (x-x_0)^{\beta_k}=
541: \sum^{\infty}_{k=0} e_k (x-x_0)^{\beta_k-\alpha/2} ,
542: \ee
543: where 
544: \be
545: \beta_k=\frac{k}{2n} .
546: \ee
547: In this case, the action of fractional derivative of order $\alpha=m/n$ 
548: can be represented as 
549: the change of the number of term $e_k \rightarrow e_{k-2m}$:
550: \be
551: D^{\alpha}_x (x-x_0)^{\beta_k} =
552: \frac{\Gamma(\beta_k+1)}{\Gamma(\beta_{k-2m}+1)} (x-x_0)^{\beta_{k-2m}} .
553: \ee
554: It allows us to derive the generalized psi-series solutions
555: of fractional Ginzburg-Landau equation.
556: 
557: Substitution of the series
558: \be
559: Z(x)=\sum^{\infty}_{k=0} e_k (x-x_0)^{\beta_k-\alpha/2}=
560: \sum^{\infty}_{k=0} e_k (x-x_0)^{\frac{k-m}{2n}}
561: \ee
562: into equation (\ref{FGL}) gives
563: \[ g \sum^{\infty}_{k=0} e_k \frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})}
564: (x-x_0)^{\frac{k-3m}{2n}}
565: +c \sum^{\infty}_{k=0} e_k \frac{k-m}{2n} (x-x_0)^{\frac{k-m-2n}{2n}}+ \]
566: \be \label{ser1}
567:  +a \sum^{\infty}_{k=0} e_k (x-x_0)^{\frac{k-m}{2n}}+
568: b \sum^{\infty}_{k_1=0} \sum^{\infty}_{k_2=0}\sum^{\infty}_{k_3=0} 
569: e_{k_1} e_{k_2} e_{k_3}  (x-x_0)^{\frac{k_1+k_2+k_3-3m}{2n}}=0. \ee
570: Let us compute $e_k$ ($k=1,2,...$) through the equation
571: of coefficients of like power of $(x-x_0)$ to zero in (\ref{ser1}):
572: \[ g \sum^{\infty}_{k=0} e_k 
573: \frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})}
574: (x-x_0)^{\frac{k-3m}{2n}}
575: +c \sum^{\infty}_{k=2m-2n} e_{k-2m+2n} 
576: \frac{k-3m+2n}{2n} (x-x_0)^{\frac{k-3m}{2n}}+ \]
577: \[ +a \sum^{\infty}_{k=2m} e_{k-2m} (x-x_0)^{\frac{k-3m}{2n}}
578: +3b e^2_0 \sum^{\infty}_{k=0} e_k (x-x_0)^{\frac{k-3m}{2n}} +\]
579: \be \label{ser2} 
580: +b \sum^{\infty}_{k=0} \sum^{k-1}_{i=1} \sum^{i}_{j=1} 
581: e_{k-i-j} e_{i} e_{j}  (x-x_0)^{\frac{k-3m}{2n}}=0. \ee
582: Using $e^2_0=f^2$, we get
583: \[ g e_k \frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})}
584: +c e_{k-2m+2n} \frac{k-3m+2n}{2n} + \]
585: \be \label{ser3}
586: +a e_{k-2m} +3b f^2 e_k +
587: b \sum^{k-1}_{i=1} \sum^{i}_{j=1} e_{k-i-j} e_{i} e_{j} =0. \ee
588: Here $k=0,1,2,...$, and $e_k=0$ for $k<0$.
589: We can rewrite the recurrence relation (\ref{ser3}) as
590: \be \label{r1} 
591: e_k \left(g\frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})}
592: + 3b f^2 \right)=
593: -c e_{k-2m+2n} \frac{k-3m+2n}{2n} -a e_{k-2m} 
594: -b \sum^{k-1}_{i=1} \sum^{i}_{j=1} e_{k-i-j} e_{i} e_{j} , \ee
595: where 
596: \be \label{f2}
597: f^2=-\frac{g\Gamma(1-\alpha/2)}{b\Gamma(1-3\alpha/2)}=
598: -\frac{g\Gamma(\frac{2n-m}{2n})}{b\Gamma(\frac{2n-3m}{2n})} .
599: \ee
600: Substitution of equation (\ref{f2}) into equation (\ref{r1}) gives 
601: \[ e_k g \left(\frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})}
602: -\frac{3\Gamma(\frac{2n-m}{2n})}{\Gamma(\frac{2n-3m}{2n})} \right)=
603: -c e_{k-2m+2n} \frac{k-3m+2n}{2n} -a e_{k-2m}- \]
604: \be
605: -b \sum^{k-1}_{i=1} \sum^{i}_{j=1} e_{k-i-j} e_{i} e_{j} . \ee
606: Note that we get resonances for the $k$ that satisfies the condition
607: \be
608: \frac{\Gamma(\frac{k-m+2n}{2n})}{\Gamma(\frac{k-3m+2n}{2n})} 
609: -\frac{3\Gamma(\frac{2n-m}{2n})}{\Gamma(\frac{2n-3m}{2n})}=0. 
610: \ee
611: In this case, the coefficient $e_k$ can be arbitrary.
612: 
613: As a result,  we obtain the nonresonance terms
614: \be \label{mrec}
615: e_k= -A(k,m,n)
616: \left(c e_{k-2m+2n} \frac{k-3m+2n}{2n} +a e_{k-2m} +b 
617: \sum^{k-1}_{i=1} \sum^{i}_{j=1} e_{k-i-j} e_{i} e_{j} \right) ,
618: \ee
619: where
620: \be \label{mrec2}
621: A(k,m,n)= - \frac{ \Gamma(\frac{k-3m+2n}{2n}) \Gamma(\frac{2n-3m}{2n}) }{g 
622: [ \Gamma(\frac{k-m+2n}{2n})\Gamma(\frac{2n-3m}{2n}) 
623: -3 \Gamma(\frac{k-3m+2n}{2n}) \Gamma(\frac{2n-m}{2n}) ]} .
624: \ee
625: 
626: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
627: \section{Fractional Ginzburg-Landau Equation with $\alpha=1.5$}
628: 
629: Let us consider the FGL equation (\ref{FGL}) with derivative
630: of order $\alpha=3/2$.
631: In this case, $n=2$, $m=3$, 
632: and the coefficients $e_k$ are defined by 
633: \be \label{e_k}
634: e_k= -\frac{\left(c e_{k-2} \frac{k-5}{4} +a e_{k-6} +b 
635: \sum^{k-1}_{i=1} \sum^{i}_{j=1} e_{k-i-j} e_{i} e_{j}\right) 
636: \Gamma(\frac{k-5}{4}) \Gamma(\frac{-5}{4}) }{
637: g [ \Gamma(\frac{k+1}{4})\Gamma(\frac{-5}{4}) 
638: -3 \Gamma(\frac{k-5}{4}) \Gamma(\frac{1}{4})]} . \ee
639: 
640: For $k=0$, we have
641: \be \label{e0}
642: e_0=f=\sqrt{ -\frac{g\Gamma(\frac{1}{4})}{b\Gamma(\frac{-5}{4})} } ,
643: \ee
644: where $\Gamma(\frac{1}{4})/\Gamma(\frac{-5}{4})>0$, and 
645: we suppose $g/b<0$.
646: For $k=1$, equation (\ref{e_k}) leads to $e_1=0$.
647: For $k=2$,
648: \be \label{k=2}
649: e_2= \frac{3 c e_{0}   
650: \Gamma(\frac{-3}{4}) \Gamma(\frac{-5}{4}) }{
651: 4 g [ \Gamma(\frac{3}{4})\Gamma(\frac{-5}{4})
652: -3 \Gamma(\frac{-3}{4}) \Gamma(\frac{1}{4})]} . \ee
653: Using the relation (\ref{e0}), we get 
654: \be \label{e2}
655: e_2=-\frac{c \sqrt{-5(g/b)} \pi^{3/2} 2^{3/4} \Gamma(3/4) }{2g 
656: [2 \Gamma^4(3/4)+5\pi^2]} .
657: \ee
658: For $k=3$, $e_3=0$. For $k=4$, 
659: \be \label{k=4}
660: e_4= -\frac{\left(c e_{2} \frac{-1}{4}+b e_{0} e^2_{2} \right)
661: \Gamma(\frac{-1}{4}) \Gamma(\frac{-5}{4}) }{g 
662: [ \Gamma(\frac{5}{4})\Gamma(\frac{-5}{4}) 
663: -3 \Gamma(\frac{-1}{4}) \Gamma(\frac{1}{4}) ]} . \ee
664: Substitution of (\ref{e0}), and (\ref{e2}) into (\ref{k=4}) gives 
665: \be \label{e4}
666: e_4=\frac{c^2 \sqrt{5g\pi \sqrt{2}/b} 
667: \Gamma^7(3/4)}{4g^2[2\Gamma^4(3/4)+5\pi^2]^2} 
668: \ee
669: For $k=5$, we have $e_5=0$. For $k=6$, equation (\ref{e_k}) is
670: \be \label{k=6}
671: e_6= -\frac{\left(c e_{4} \frac{1}{4} +
672: a e_{0} +b e_2^3 +2b e_0 e_2 e_4 \right) 
673: \Gamma(\frac{1}{4}) \Gamma(\frac{-5}{4}) }{g 
674: [ \Gamma(\frac{7}{4})\Gamma(\frac{-5}{4}) 
675: -3 \Gamma(\frac{1}{4}) \Gamma(\frac{1}{4})]} . \ee
676: Substituting $e_0$, $e_2$ and $e_4$ from 
677: equations (\ref{e0}), (\ref{e2}), and (\ref{e4}) into (\ref{k=6}), 
678: we have
679: \[ e_6=-\frac{\sqrt{-5g/b} \pi^{3/2} 2^{3/4} \Gamma(3/4) }{3g^2 
680: [2\Gamma^4(3/4)+5\pi^2]^3 [2 \Gamma^4(3/4)-5\pi^2]} \Bigl(
681: c^3\Gamma^{12}(3/4)+5c^3\Gamma^8(3/4) \pi^2+ \]
682: \be \label{e6}
683: +20c^3 \Gamma^4(3/4) \pi^4+16a g^2 \Gamma^{12}(3/4) 
684: +120ag^2\Gamma^{8}(3/4) \pi^2 +300ag^2\Gamma^{4}(3/4) 
685: \pi^4+250ag^2\pi^6 \Bigr) . 
686: \ee
687: 
688: 
689: As the result, we obtain 
690: \[ Z(x)=\sum^{\infty}_{k=0} e_k (x-x_0)^{\frac{k-3}{4}}= \]
691: \be \label{coe}
692: =e_0(x-x_0)^{-3/4}+e_2(x-x_0)^{-1/4}+e_4(x-x_0)^{1/4}+e_6(x-x_0)^{3/4}+...
693: \ee
694: This is a power psi-series that present the solution of 
695: FGL equation of order $\alpha=1.5$.
696: The coefficients in equation (\ref{coe}) are defined by equations 
697: (\ref{e0}), (\ref{e2}), (\ref{e4}), and (\ref{e6}).
698: For example, $a=-b=c=g=1$ gives
699: \[ e_0 \approx 0.9615539375 , \quad e_2 \approx -0.2382246293, \] 
700: \[ e_4 \approx 0.001685563496 ,\quad e_6 \approx 0.3872134448 .\]
701: 
702: 
703: 
704: 
705: \section{Next to Singular Behavior}
706: 
707: In sections 5 and 6, we consider the rational $\alpha$. 
708: In this section, we suppose that the order $\alpha$ 
709: is an arbitrary (rational or irrational). 
710: Instead of imposing a series commencing at the power 
711: indicated by the singularity found by the leading-order analysis, 
712: we can determine the next to singular behavior by writing
713: \be \label{ZfpG}
714: Z(x)=f (x-x_0)^p +G(x), 
715: \ee
716: where
717: \be \label{p-f}
718: p=-\frac{\alpha}{2}, \quad
719: f=\sqrt{-\frac{g\Gamma(1-\alpha/2)}{b\Gamma(1-3\alpha/2)}} .
720: \ee
721: We can always write $Z(x)$ in the form (\ref{ZfpG}). 
722: To make the process useful, we require that
723: the first term be the leading order term, i.e., 
724: \be \label{xG0}
725: (x-x_0)^{-p}G(x)=(x-x_0)^{\alpha/2}G(x) \rightarrow 0 \quad
726: if \quad (x-x_0) \rightarrow 0 . 
727: \ee 
728: Substituting (\ref{ZfpG}) into (\ref{FGL}), and using (\ref{Da}),
729: we have
730: \[ gf \frac{\Gamma(p+1)}{\Gamma(p+1-\alpha)} (x-x_0)^{p-\alpha}+
731: cpf(x-x_0)^{p-1}+af (x-x_0)^p+ bf^3 (x-x_0)^{3p}+\]
732: \be \label{66}
733: +gD^{\alpha}_x G(x)+cD^1_x G(x)+aG(x)+bG^3(x)
734: +3 bf^2 (x-x_0)^{2p} G(x)+ 3bf (x-x_0)^{p} G^2(x)=0 .
735: \ee
736: Equations (\ref{66}) and (\ref{p-f}) give
737: \[ gD^{\alpha}_x G(x)+cD^1_x G(x)+aG(x)+bG^3(x)
738: +3 bf^2 (x-x_0)^{2p} G(x)+ \]
739: \be
740: +3bf (x-x_0)^{p} G^2(x) cpf(x-x_0)^{p-1}+af (x-x_0)^p=0 .
741: \ee
742: Multiplying this equation on $(x-x_0)^{-3p}$,
743: and using condition (\ref{xG0}), we get
744: the equation without nonlinear terms for $(x-x_0)\rightarrow 0$.
745: 
746: As the result, we obtain
747: \be \label{linG}
748: gD^{\alpha}_x G(x)+cD^1_x G(x)+aG(x)+
749: cpf(x-x_0)^{-\alpha/2-1}+af (x-x_0)^{-\alpha/2}=0 ,
750: \ee
751: with condition (\ref{xG0}) for the solutions.
752: Equation (\ref{linG}) is a linear inhomogeneous fractional equation.
753: The solution of this equation allows us to find the
754: solution of the one-dimensional FGL equation.
755: 
756: Let us consider equation (\ref{linG}) with $c=0$, and the 
757: boundary conditions
758: \be
759: \left(D^{\alpha-k}_x G(x)\right)_{x=x_0}=G_k, \quad k=1, 2.
760: \ee
761: Then the solution is
762: \[ G(x)=\sum^{2}_{k=1} G_k (x-x_0)^{\alpha-k} 
763: E_{\alpha,\alpha+1-k} [-a (x-x_0)^{\alpha}]+ \]
764: \be \label{Gsol1}
765: +af \int^{x}_0 (x-x_0-y)^{\alpha-1} E_{\alpha, \alpha} [-a(x-x_0-y)^{\alpha}]
766: (y-x_0)^{-\alpha/2} dy .
767: \ee
768: Here $E_{\alpha,\beta}[z]$ is a Mitag-Leffler function that is defined by
769: \be
770: E_{\alpha,\beta}[z]=\sum^{\infty}_{k=0} \frac{z^k}{\Gamma(\alpha k+\beta)}.
771: \ee
772: Let us consider the integral representation 
773: for the Mittag-Leffler function
774: \be \label{58}
775: E_{\alpha,\beta}[z]=\frac{1}{2\pi i}\int_{Ha} 
776: \frac{\xi^{\alpha-\beta} e^{\xi}}{\xi^{\alpha}-z} d\xi ,
777: \ee
778: where $Ha$ denotes the Hankel path, 
779: a loop which starts from $-\infty$ along the lower side
780: of the negative real axis, encircles the circular disc 
781: $|\xi|\le |z|^{1/\alpha}$ in the positive direction, and 
782: ends at $-\infty$ along the upper side of the negative real axis.
783: By the replacement $\xi^{\alpha} \rightarrow \xi$ equation (\ref{58})
784: transforms into \cite{Podlubny,GLL}:
785: \be \label{59}
786: E_{\alpha,\beta}[z]=\frac{1}{2 \pi i \alpha} 
787: \int_{\gamma(a,\delta)} \frac{e^{ \xi^{1/\alpha} } 
788: \xi^{(1-\beta)/\alpha}}{\xi -z} d\xi, \quad (1<\alpha <2) ,
789: \ee
790: where $\pi \alpha/2 < \delta < min\{\pi, \pi \alpha\}$.
791: The contour $\gamma(a,\delta)$ consists of two rays
792: $S_{-\delta}=\{\arg(\xi)=-\delta, |\xi|\ge a\}$ and  
793: $S_{+\delta}=\{\arg(\xi)=+\delta, |\xi|\ge a\}$,
794: and a circular arc 
795: $C_{\delta}=\{|\xi|=1, -\delta \le arc(\xi) \le \delta \}$.
796: Let us denote the region on the left from $\gamma(a,\delta)$
797: as $G^{-}(a,\delta)$. Then \cite{GLL}:
798: \be \label{60}
799: E_{\alpha,\beta}[z]=-\sum^{\infty}_{n=1} 
800: \frac{z^{-n}}{\Gamma(\beta-\alpha n)}, \quad z \in G^{-}(a,\delta), 
801: \quad (|z| \rightarrow \infty),  
802: \ee 
803: and $\delta \le |\arg(z)|\le \pi$.
804: In our case, $z=-a(x-x_0)^{\alpha}$, $\arg(z)=\pi$.
805: As a result, we arrive at the asymptotic of the solution, 
806: which exhibits power-like tails for $x \rightarrow \infty$.
807: These power-like tails are the most important effect of 
808: the noninteger derivative in the fractional equations.
809: 
810: 
811: \section{Conclusion}
812: 
813: 
814: In this paper, we generalize the
815: psi-series approach to the fractional differential equations.
816: As an example, we consider the fractional Ginzburg-Landau (FGL) 
817: equation \cite{Zaslavsky6,Physica2005,Mil,TZ2}.
818: The suggested psi-series approach can be used for 
819: a wide class of fractional nonlinear equations. 
820: The leading-order behaviours of solutions about an arbitrary 
821: singularity, as well as their resonance structures, 
822: can be derived for fractional equations by the suggested 
823: generalization of psi-series.
824: 
825: In the paper, we use the series
826: \be \label{Con1}
827: Z(x)=\frac{1}{(x-x_0)^{m/2n}}
828: \sum^{\infty}_{k=0} e_k \; (x-x_0)^{k/2n} ,
829: \ee
830: where $k$, $m$, $n$ are the integer numbers. 
831: For the order $\alpha=m/n$, the action of fractional derivative 
832: \be
833: D^{\alpha}_x (x-x_0)^{k/2n} =
834: \frac{\Gamma(k/2n+1)}{\Gamma((k-2m)/2n+1)} (x-x_0)^{(k-2m)/2n} 
835: \ee
836: can be represented as 
837: the change of the number of term $e_k \rightarrow e_{k-2m}$ in (\ref{Con1}).
838: It allows us to derive the psi-series 
839: for the fractional differential equation of order $\alpha=m/n$.
840: For the FGL equation the leading-order singular behaviour
841: is defined by power that is equal to the half of derivative order. 
842: 
843: A remarkable property of the dynamics described by the equation with 
844: fractional space derivatives is that the 
845: solutions have power-like tails \cite{KZT}. 
846: In this paper, we prove that fractional differential 
847: equations of order $\alpha$ with a polynomial nonlinear 
848: term of order $s$ have the noninteger power-like behavior 
849: of order $\alpha/(1-s)$ near the singularity.
850: 
851: It is interesting to find barriers to integrability 
852: for fractional differential equations. 
853: In general, the integrability of fractional nonlinear equations 
854: is a very interesting object for future researches. 
855: It has many problems that are connected with specific 
856: properties of the fractional calculus.
857: For example, we must derive a generalization 
858: of the Lie algebra for the vector fields that are
859: defined by fractional derivatives.
860: For this generalization, the Jacobi identity cannot be satisfied, 
861: and we have a non-Lie algebra. 
862: The definition of such "fractional" Lie algebra is an open question 
863: and cannot be realized by a simple way. 
864: To formulate the fractional generalization of a Lie algebra 
865: for derivatives of noninteger oder, 
866: we can use the representation of fractional derivatives 
867: as infinite series of derivatives of integer orders \cite{SKM}.
868: For example, the Riemann-Liouville fractional derivative 
869: can be represented as 
870: \be
871: D^{\alpha}_x=\sum^{\infty}_{n=0} A_n(x,x_0,\alpha)
872: \frac{d^n}{dx^n} ,
873: \ee
874: where
875: \be
876: A_n(x,a,\alpha)= 
877: \frac{(-1)^{n-1}\alpha \Gamma(n-\alpha)}{\Gamma(1-\alpha) \Gamma(n+1)}
878: \frac{(x-x_0)^{n-\alpha}}{\Gamma(n+1-\alpha)}  .
879: \ee
880: %%%for the functions that are analytical in the interval $(x_0,x_1)$.
881: Then the possible realization of the generalization is connected with 
882: the special algebraic structures for infinite jets \cite{Vin}.
883: These structures and approaches can help 
884: to solve some problems that are connected with
885: the integrability of fractional nonlinear equations. 
886: 
887: 
888: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
889: \begin{thebibliography}{**}
890: \small
891: 
892: \bi{SKM} S.G. Samko, A.A. Kilbas, O.I. Marichev,
893: {\it Fractional Integrals and Derivatives Theory and Applications}
894: (Gordon and Breach, New York, 1993)
895: 
896: \bi{OS} K.B. Oldham, J. Spanier,
897: {\it The Fractional Calculus} 
898: (Academic Press, New York, 1974)
899: 
900: \bi{Podlubny} I. Podlubny, {\it Fractional Differential Equations}
901: (Academic Press, New York, 1999) 
902: 
903: \bi{KST} A.A. Kilbas, H.M. Srivastava, J.J. Trujillo,
904: {\it Theory and Applications of Fractional Differential Equations} 
905: (Elsevier, Amsterdam, 2006) 
906: 
907: \bi{Zaslavsky1} G.M. Zaslavsky, 
908: "Chaos, fractional kinetics, and anomalous transport"
909: Phys. Rep. {\bf 371} (2002) 461-580.
910: 
911: \bi{Zaslavsky2}  G.M. Zaslavsky, 
912: {\it Hamiltonian Chaos and Fractional Dynamics} 
913: (Oxford University Press, Oxford, 2005)
914: 
915: \bi{Hilfer} {\it Applications of Fractional Calculus in Physics} 
916: Ed. by R. Hilfer (World Scientific, Singapore, 2000)
917: 
918: \bi{C2} M. Caputo, {\it Elasticita e Dissipazione} 
919: (Zanichelli, Bologna, 1969)
920: 
921: \bi{Nig1} R.R. Nigmatullin,
922: "The realization of the generalized transfer equation
923: in a medium with fractal geometry"
924: Phys. Status Solidi B {\bf 133} (1986) 425-430;
925: "Fractional integral and its physical interpretation"
926: Theor. Math. Phys. {\bf 90} (1992) 242-251. 
927: 
928: \bi{Mainardi} A. Carpinteri, F. Mainardi, 
929: {\it Fractals and Fractional Calculus in Continuum Mechanics} 
930: (Springer, New York, 1997)
931: 
932: \bi{Media} V.E. Tarasov,
933: "Continuous medium model for fractal media"
934: Phys. Lett. A {\bf 336} (2005) 167-174 (cond-mat/0506137);
935: "Fractional hydrodynamic equations for fractal media" 
936: Ann. Phys. {\bf 318} (2005) 286-307 (physics/0602096);
937: "Fractional Fokker-Planck equation for fractal media"
938: Chaos {\bf 15} (2005) 023102 (nlin.CD/0602029);
939: "Wave equation for fractal solid string"
940: Mod. Phys. Lett. B {\bf 19} (2005) 721-728 (physics/0605006);
941: "Possible experimental test of continuous 
942: medium model for fractal media"
943: Phys. Lett. A {\bf 341} (2005) 467-472 (physics/0602121).
944: 
945: 
946: \bi{Laskin} N. Laskin,
947: "Fractals and quantum mechanics"  
948: Chaos {\bf 10} (2000) 780-790;
949: "Fractional quantum mechanics" 
950: Phys. Rev. E {\bf 62} (2000) 3135-3145;
951: "Fractional quantum mechanics and Levy path integrals"
952: Phys. Lett. A {\bf 268} (2000) 298-305 (hep-ph/9910419);
953: "Fractional Schrodinger equation"
954: Phys. Rev. E {\bf 66} (2002) 056108 (quant-ph/0206098).
955: 
956: \bi{Naber} M. Naber,
957: "Time fractional Schrodinger equation"
958: J. Math. Phys. {\bf 45} (2004) 3339-3352 (math-ph/0410028).
959: 
960: \bi{Zaslavsky7} G.M. Zaslavsky, 
961: "Fractional kinetic equation for Hamiltonian chaos"
962: Physica D {\bf 76} (1994) 110-122.
963: 
964: \bi{SZ} A.I. Saichev, G.M. Zaslavsky,
965: "Fractional kinetic equations: solutions and applications"
966: Chaos {\bf 7} (1997) 753-764.
967: 
968: \bi{ZE} G.M. Zaslavsky, M.A. Edelman,
969: "Fractional kinetics: from pseudochaotic 
970: dynamics to Maxwell's demon"
971: Physica D {\bf 193} (2004) 128-147.
972: 
973: \bi{CLZ} B.A. Carreras, V.E. Lynch, G.M. Zaslavsky,
974: "Anomalous diffusion and exit time distribution of 
975: particle tracers in plasma turbulence model"
976: Physics of Plasmas {\bf 8} (2001) 5096-5103.
977: 
978: \bi{Plasma2005} V.E. Tarasov,
979: "Electromagnetic field of fractal distribution of charged particles"
980: Physics of Plasmas {\bf 12} (2005) 082106;
981: "Multipole moments of fractal distribution of charges"
982: Mod. Phys. Lett. B. {\bf 19} (2005) 1107-1118;
983: "Magnetohydrodynamics of fractal media"
984: Physics of Plasmas {\bf 13} (2006) 052107.
985: 
986: \bi{CMDA} V.E. Tarasov,
987: "Gravitational field of fractal distribution of particles" 
988: Celest. Mech. Dynam. Astron. {\bf 19} (2006) 1-15 (astro-ph/0604491).
989: 
990: \bi{GM} F. Mainardi, R. Gorenflo,
991: "On Mittag-Leffler-type functions in 
992: fractional evolution processes"
993: J. Comput. Appl. Math. {\bf 118} (2000) 283-299. 
994: 
995: \bi{nonHam} V.E. Tarasov, 
996: "Fractional generalization of Liouville equation,"
997: Chaos {\bf 14} (2004) 123-127 (nlin.CD/0312044);
998: "Fractional systems and fractional Bogoliubov hierarchy equations"
999: Phys. Rev. E {\bf 71} (2005) 011102 (cond-mat/0505720);
1000: "Fractional Liouville and BBGKI equations" 
1001: J. Phys. Conf. Ser. {\bf 7} (2005) 17-33 (nlin.CD/0602062);
1002: "Transport equations from Liouville equations for fractional systems"
1003: Int. J. Mod. Phys. B. {\bf 20} (2006) 341-353 (cond-mat/0604058).
1004: 
1005: \bi{FracHam} V.E. Tarasov,
1006: "Fractional generalization of gradient and Hamiltonian systems" 
1007: J. Phys. A  {\bf 38} (2005) 5929-5943 (math.DS/0602208).
1008: 
1009: \bi{LZ} N. Laskin, G.M. Zaslavsky, 
1010: "Nonlinear fractional dynamics on a lattice with long range interactions" 
1011: Physica A {\bf 368} (2006) 38-54 (nlin.SI/0512010).
1012: 
1013: \bi{TZ3}  V.E. Tarasov, G.M. Zaslavsky,
1014: "Fractional dynamics of coupled oscillators with long-range interaction"
1015: Chaos {\bf 16} (2006) 023110 (nlin.PS/0512013);
1016: "Fractional dynamics of systems with long-range interaction"
1017: Commun. Nonlin. Sci. Numer. Simul. {\bf 11} (2006) 885-898.
1018: 
1019: \bi{KZT}  N. Korabel, G.M. Zaslavsky, V.E. Tarasov,
1020: "Coupled oscillators with power-law interaction and 
1021: their fractional dynamics analogues"
1022: Commun. Nonlin. Sci. Numer. Simul. 
1023: {\bf 11} (2006) accepted (math-ph/0603074)
1024: 
1025: \bi{Montr} E.W. Montroll, M.F. Shlesinger, 
1026: "The wonderful world of random walks"
1027: In: {\it Studies in Statistical Mechanics} Vol. {\bf 11}
1028: (North-Holland, Amsterdam, 1984) pp.1-121.
1029: 
1030: \bi{Uch} V.V. Uchaikin,
1031: "Self-similar anomalous diffusion and Levy-stable laws"
1032: Physics-Uspekhi {\bf 46} (2003) 821-849;
1033: "Anomalous diffusion and fractional stable distributions" 
1034: J. Exper. Theor. Phys. {\bf 97} (2003) 810-825.
1035: 
1036: \bi{Zaslavsky6} H. Weitzner, G.M. Zaslavsky,  
1037: "Some applications of fractional derivatives"
1038: Commun. Nonlin. Sci. and Numer. Simul. 
1039: {\bf 8} (2003) 273-281 (nlin.CD/0212024).
1040: 
1041: \bi{Physica2005} V.E. Tarasov, G.M. Zaslavsky,
1042: "Fractional Ginzburg-Landau equation for fractal media"
1043: Physica A  {\bf 354} (2005) 249-261 (physics/0511144).
1044: 
1045: \bi{Mil} A.V. Milovanov, J.J. Rasmussen, 
1046: "Fractional generalization of the Ginzburg-Landau equation: 
1047: an unconventional approach to critical phenomena in complex media"
1048: Phys. Lett. A {\bf 337} (2005) 75-80 (cond-mat/0309577). 
1049: 
1050: \bi{TZ2}  V.E. Tarasov, G.M. Zaslavsky
1051: "Dynamics with low-level fractionality"
1052: Physica A {\bf 368} (2006) 399-415 (physics/0511138)
1053: 
1054: \bi{Leon} M.A. Leontovich, Izvestya Akademii Nauk SSSR, 
1055: Ser. Fiz. {\bf 8} (1944) 16.
1056: 
1057: \bi{Light} J. Lighthill, {\it Waves in Fluid} 
1058: (Cambridge Univ. Press, Cambridge, 1978) Sec. 3.7.
1059: 
1060: \bi{Kad} B.B. Kadomtsev, {\it Collective Phenomena in Plasmas} 
1061: (Pergamon, New York, 1978). Sec. 2.5.5. and 3.4.4.
1062: %%%Kadomtsev BB, “Collective Phenomena in Plasma”, 
1063: %%%2nd ed. Nauka, Moscow (1988) (in Russian).
1064: 
1065: \bi{ZS} R.Z. Sagdeev, D.A. Usikov, G.M. Zaslavsky,
1066: {\it Nonlinear Physics. From the Pendulum to Turbulence and Chaos} 
1067: (Harwood Academic, New York, 1988) Sec. 15.1.
1068: 
1069: \bi{Ince} E.L. Ince, {\it Ordinary Differential Equations}, 
1070: (Longmans-Green, London, 1927).
1071: 
1072: \bi{3} F.J. Bureau, 
1073: "Differential equations with fixed critical points" 
1074: Annali di Matematica pura e applicata {\bf 116} (1964) 1-116.
1075: 
1076: \bi{4} C.M. Cosgrove, G. Scoufis, 
1077: "Painleve classification of a class of differential equations 
1078: of the second order and second degree", 
1079: Stud. Appl. Math. {\bf 88} (1993) 25-87.
1080: 
1081: \bi{Ablowitz} M.J. Ablowitz, A. Ramani, H. Segur, 
1082: "Nonlinear evolution equations and ordinary
1083: differential equations of Painleve type" 
1084: Lett. Nuovo Cimento {\bf 23} (1978) 333-337;
1085: "A connection between nonlinear evolution
1086: equations and ordinary differential equations of P type I" 
1087: J. Math. Phys. {\bf 21} (1980) 715-721;
1088: "A connection between nonlinear evolution
1089: equations and ordinary differential equations of P type II" 
1090: J. Math. Phys. {\bf 21} (1980) 1006-1015.
1091: 
1092: \bi{8} R. Conte, 
1093: "Singularities of differential equations and integrability" 
1094: in {\it Introduction to Methods of Complex Analysis and 
1095: Geometry for Classical Mechanics and Nonlinear Waves}, 
1096: Editors: Benest D and Froeschle C (Editions Frontieres, 
1097: Gif-sur-Yvette, 1993)
1098: 
1099: \bi{Tabor1} M. Tabor, 
1100: {\it Chaos and Integrability in Nonlinear Dynamics}
1101: (Wiley, New York, 1989) Sec. 8.
1102:  
1103: \bi{Tabor2} M. Tabor, J. Weiss,
1104: "Analytic structure of the Lorenz system"
1105: Phys. Rev. A {\bf 24} (1981) 2157-2167.
1106: 
1107: \bi{BSV} T. Bountis, H. Segur, F. Vivaldi, 
1108: "Integrable Hamiltonian systems and the Painleve property"
1109: Phys. Rev. A {\bf 25} (1982) 1257-1264.
1110: 
1111: \bi{CTW} Y.F. Chang, M. Tabor, J. Weiss,
1112: "Analytic structure of the Henon-Heiles Hamiltonian 
1113: in integrable and nonintegrable regime"
1114: J. Math. Phys. {\bf 23} (1982) 531.
1115: 
1116: \bi{GLL} R. Gorenflo, J. Loutchko, Y. Luchko,
1117: "Computation of the Mittag-Leffler function and its derivative"
1118: Fractional Calcul. and Appl. Anal. 
1119: {\bf 5} (2002) 491-518.
1120: 
1121: \bi{Vin}
1122: A.M. Vinogradov, I.S. Krasil'shchik, V.V. Lychagin, 
1123: {\it Introduction to the Geometry of Nonlinear Differential Equations} 
1124: (Nauka, Moscow, 1986) in Russian;
1125: "Algebraic aspects of differential calculus" 
1126: (a collection of papers, edited by  
1127: A.M. Vinogradov and I.S. Krasil'shchik)
1128: Acta Applicandae Mathematicae,  
1129: {\bf 49} Special Issue 3, (1997);
1130: A.M. Vinogradov, M.M. Vinogradov, 
1131: "Graded multiple analogs of Lie algebras" 
1132: Acta Appl. Math. {\bf 72} (2002) 183-197.
1133: 
1134: 
1135: 
1136: 
1137: \end{thebibliography}
1138: 
1139: 
1140: 
1141: \end{document}
1142: 
1143: 
1144: