arXiv reader

Discrete Painlevé transcendent solutions to the multiplicative type discrete KdV equations

Nobutaka Nakazono

arXiv
2104.11433v2 · PDF · 9 March 2022
Categories
nlin.SI, math-ph, math.MP
DOI
10.1063/5.0054905
Licence
CC BY 4.0 The site shows the text.

Abstract

Hirota’s discrete KdV equation is an integrable partial difference equation on \mathbb{Z}^2, which approaches the Korteweg-de Vries (KdV) equation in a continuum limit. In this paper, we show that its multiplicative-discrete versions have the special solutions given by the solutions of q-Painlevé equations of types A_J^{(1)} (J=3,4,5,6).

Items 11

Remark

Remark 1.1

YXHO

As multiplicative-discrete type difference equations are also called q-difference equations, it is common to use the parameter q for their shift parameters. However, in this paper, the parameter q is used to denote the shift parameters of q-Painlevé equations, and we also consider the correspondence between the shift parameters of the q-Painlevé equations and those of the multiplicative dKdV equations. Therefore, we use the parameter \epsilon instead of the parameter q for the multiplicative dKdV equations (?) and (?) to avoid confusion.

Theorem

Theorem 1.2 (A_6^{(1)}-type)

SLNZ

The multiplicative dKdV equation (?) has the special solution

u_{l,m}=\dfrac{\epsilon^{3/4}{\beta_{m}}^{1/2}(\alpha_{l}-\beta_{m})}{{\alpha_{l}}^{1/2}(h_{l,m}+\beta_{m+2})g_{l,m}}.

Here, the functions g_{l,m} and h_{l,m} satisfy the system of ordinary difference equations for the l-direction:

h_{l+1,m}h_{l,m}=\dfrac{\beta_{m+2}(1+g_{l,m})}{{g_{l,m}}^2},\quad g_{l+1,m}g_{l,m}=\dfrac{\beta_{m}(h_{l+1,m}+\alpha_{l+3})}{\alpha_{l+1}h_{l+1,m}(h_{l+1,m}+\beta_{m+2})},

and that for the m-direction:

f_{l,m+1}f_{l,m}=\dfrac{\alpha_l (h_{l,m}+\beta_{m+2})}{\beta_mh_{l,m}(h_{l,m}+\alpha_{l+2})},\quad h_{l,m+1}h_{l,m}=\dfrac{\alpha_{l+2}(1+f_{l,m+1})}{{f_{l,m+1}}^2},

where

f_{l,m}=\dfrac{1}{g_{l,m}h_{l,m}}.

Note that each of equations (1.2) and (1.2) is equivalent to the q-P_{\rm III}^{D_7^{(1)}} (?), which is a q-Painlevé equation of A_6^{(1)}-type. (See Appendix ? for details.)

Theorem

Theorem 1.3 (A_5^{(1)}-type)

2LAX

The multiplicative dKdV equation (?) has the special solution

u_{l,m}=\dfrac{d(\alpha_{l}-\beta_{m})}{({\alpha_{l}}^{1/2}+{\beta_{m}}^{1/2}g_{l,m})\,h_{l,m}},

where d\in\mathbb{C}^\ast is an arbitrary parameter. Here, the functions g_{l,m} and h_{l,m} satisfy the system of ordinary difference equations for the l-direction:

g_{l+1,m}g_{l,m} =\dfrac{d^2({\alpha_{l}}^{1/2}+f_{l,m})}{f_{l,m}(1+{\alpha_{l}}^{1/2}f_{l,m})},\quad f_{l+1,m}f_{l,m} =\dfrac{d^2({\alpha_{l+1}}^{1/2}+{\beta_{m}}^{1/2}g_{l+1,m})}{g_{l+1,m}({\beta_{m}}^{1/2}+{\alpha_{l+1}}^{1/2}g_{l+1,m})},

where

f_{l,m}=\dfrac{d^2}{g_{l,m}h_{l,m}},

and that for the m-direction:

h_{l,m+1}h_{l,m} =\dfrac{d^2({\beta_{m}}^{1/2}+{\alpha_{l}}^{1/2}g_{l,m})}{g_{l,m}({\alpha_{l}}^{1/2}+{\beta_{m}}^{1/2}g_{l,m})},\quad g_{l,m+1}g_{l,m} =\dfrac{d^2({\beta_{m}}^{1/2}+h_{l,m+1})}{h_{l,m+1}(1+{\beta_{m}}^{1/2}h_{l,m+1})}.

Note that each of equations (1.3) and (1.3) is equivalent to the q-P_{\rm III} (?), which is a q-Painlevé equation of A_5^{(1)}-type. (See Appendix ? for details.)

Theorem

Theorem 1.4 (A_4^{(1)}-type)

G85A

The multiplicative dKdV equation (?) has the special solution

u_{l,m}=\dfrac{{\beta_{m}}^{1/2}(\alpha_{l}-\beta_{m})h_{l,m}}{\epsilon^{1/4}{\alpha_{l}}^{3/2}d_1d_2(1+\beta_{m} g_{l,m})},

where d_1,d_2\in\mathbb{C}^\ast are arbitrary parameters. Here, the functions g_{l,m} and h_{l,m} satisfy the system of ordinary difference equations for the l-direction:

\begin{cases} g_{l+1,m}g_{l,m}=\dfrac{(\beta_{m-1}{d_1}^2{d_2}^2+\alpha_lf_{l,m})(\beta_m+\alpha_lf_{l,m})}{{\alpha_l}^2{\beta_m}^2{d_2}^2({d_1}^2\beta_m+f_{l,m})},\\[1em] f_{l+1,m}f_{l,m}=\dfrac{{\beta_{m}}^2{d_1}^2{d_2}^2(1+\alpha_{l+1}g_{l+1,m})(1+\alpha_{l+1}\,\beta_m\, g_{l+1,m})}{{\alpha_{l+1}}^2(1+\beta_m g_{l+1,m})}, \end{cases}

where

f_{l,m}=\dfrac{{d_1}^2{d_2}^2(1+\alpha_l\,\beta_m\,g_{l,m})}{h_{l,m}},

and that for the m-direction:

\begin{cases} h_{l,m+1}h_{l,m}=\dfrac{{\alpha_l}^2{d_1}^2{d_2}^2(1+\beta_m\,g_{l,m})(1+\alpha_l\,\beta_m\,g_{l,m})}{{\beta_m}^2(1+\alpha_l\,g_{l,m})},\\[1em] g_{l,m+1}g_{l,m}=\dfrac{(\alpha_{l+1}+\beta_{m+1}h_{l,m+1})(\alpha_l\,{d_1}^2{d_2}^2+\beta_{m+1}h_{l,m+1})}{{\alpha_l}^2\,{\beta_{m+1}}^2{d_2}^2(\alpha_l {d_1}^2+h_{l,m+1})}. \end{cases}

Note that each of equations (1.4) and (1.4) is equivalent to the q-P_{\rm V} (?), which is a q-Painlevé equation of A_4^{(1)}-type. (See Appendix ? for details.)

Theorem

Theorem 1.5 (A_3^{(1)}-type)

EHS8

The multiplicative dKdV equation (?) has the special solution

u_{l,m} =\cfrac{\epsilon^{1/4}(\beta_m\gamma-\alpha_l)\left(f_{l,m}(d_1+x_{l,m})+{d_2}^{1/2}(1+d_1{d_2}^{-1}\,\beta_mx_{l,m})\right)} {{\alpha_l}^{1/2}{\beta_m}^{1/2}\gamma^{1/2}x_{l,m}\left(\beta_mf_{l,m}(d_1+x_{l,m})+{d_2}^{1/2}(1+d_1{d_2}^{-1}\,\beta_m x_{l,m})\right)},

where d_1,d_2\in\mathbb{C}^\ast are arbitrary parameters. Here, the functions f_{l,m} and x_{l,m} satisfy the system of ordinary difference equations for the l-direction:

\begin{cases} g_{l+1,m}g_{l,m}=\dfrac{(f_{l,m}+{d_1}^{-1}{d_2}^{1/2}{\alpha_l}^{-1}\gamma)(f_{l,m}+d_1{d_2}^{-1/2}\,{\alpha_l}^{-1})}{(f_{l,m}+{d_1}^{-1}{d_2}^{1/2})(f_{l,m}+d_1\,{d_2}^{-1/2})},\\[1em] f_{l+1,m}f_{l,m}=\dfrac{(g_{l+1,m}+{d_2}^{-1/2}{\alpha_l}^{-1}{\beta_{m-1}}^{1/2}\gamma^{1/2})(g_{l+1,m}+{d_2}^{1/2}{\alpha_l}^{-1}{\beta_{m+1}}^{-1/2}\gamma^{1/2})}{(g_{l+1,m}+{d_2}^{-1/2}{\beta_{m-1}}^{-1/2}\gamma^{-1/2})(g_{l+1,m}+{d_2}^{1/2}{\beta_{m-1}}^{1/2}\gamma^{1/2})}, \end{cases}

where

g_{l,m}=\dfrac{{d_2}^{1/2}\gamma+d_1\,\alpha_l\,f_{l,m}}{\alpha_l\,{\beta_{m-1}}^{1/2}\gamma^{1/2}(d_1+{d_2}^{1/2}f_{l,m})x_{l,m}},

and that for the m-direction:

\begin{cases} y_{l,m+1}y_{l,m}=\dfrac{(x_{l,m}+{d_1}^{-1}{d_2}\,{\beta_m}^{-1})(x_{l,m}+d_1{d_2}^{-1}\, {\beta_{m-1}}^{-1}\gamma^{-1})}{(x_{l,m}+d_1)(x_{l,m}+{d_1}^{-1})},\\[1em] x_{l,m+1}x_{l,m}=\dfrac{(y_{l,m+1}+{\alpha_l}^{-1/2}{\beta_m}^{-1})(y_{l,m+1}+{\alpha_l}^{1/2}{\beta_m}^{-1}\gamma^{-1})}{(y_{l,m+1}+{\alpha_l}^{1/2})(y_{l,m+1}+{\alpha_l}^{-1/2})}, \end{cases}

where

y_{l,m}=\dfrac{{\alpha_l}^{1/2}f_{l,m}(d_1+{d_2}\,\beta_{m-1}\,\gamma\, x_{l,m})}{{d_2}^{1/2}\beta_{m-1}\,\gamma(1+d_1\, x_{l,m})}.

Note that each of equations (1.5) and (?) is equivalent to the q-P_{\rm VI} (?), which is a q-Painlevé equation of A_3^{(1)}-type. (See Appendix ? for details.)

Remark

Remark 2.1

K79Q

We follow the convention that the variables and parameters not explicitly including in the actions listed in the equations above are the ones that remain unchanged under the action of the corresponding transformation. That is, the transformation acts as an identity on those variables and parameters.

Remark

Remark 2.2

PRL9

Equation (?) is equivalent to the q-P_{\rm II} (?), and each of equations (?) and (?) is equivalent to the q-P_{\rm III}^{D_7^{(1)}} (?). Indeed, the correspondence between Equations (?) and (?) is given by

F=f_0^{l_1,l_2,l_3},\quad \overline{F}=f_0^{l_1+1,l_2,l_3},\quad \underline{F}=f_0^{l_1-1,l_2,l_3},\quad t=q^{l_1-l_3}b,\quad c_1=q^{-l_2+l_3}a_1,

that between Equations (?) and (?) is given by

\begin{split} &F=\dfrac{1}{f_2^{l_1,l_2,l_3}},\quad G=\dfrac{1}{f_1^{l_1,l_2-1,l_3}},\quad \overline{F}=\dfrac{1}{f_2^{l_1,l_2+1,l_3}},\quad \overline{G}=\dfrac{1}{f_1^{l_1,l_2,l_3}},\\ &t=q^{-l_1+l_2}a_0b^{-1},\quad c_1=q^{-l_1+l_3+1}b^{-1}, \end{split}

and that between Equations (?) and (?) is given by

\begin{split} &F=\dfrac{1}{f_2^{l_1,l_2,l_3}},\quad G=\dfrac{1}{f_0^{l_1,l_2,l_3}},\quad \overline{F}=\dfrac{1}{f_2^{l_1,l_2,l_3+1}},\quad \overline{G}=\dfrac{1}{f_0^{l_1,l_2,l_3+1}},\\ &t=q^{-l_1+l_3+1}b^{-1},\quad c_1=q^{-l_1+l_2+1}{a_1}^{-1}b^{-1}. \end{split}
Remark

Remark 3.1

JEZX

Each of equations (?) and (?) is equivalent to the q-P_{\rm III} (?). The correspondence between Equations (?) and (?) is given by

\begin{split} &F=f_{0}^{l,m},\quad G=f_{1}^{l,m},\quad \overline{F}=f_{0}^{l+1,m},\quad \overline{G}=f_{1}^{l+1,m},\\ &t=q^la_0,\quad c_1=qc^2,\quad c_2=q^{-m}a_2, \end{split}

while that between Equations (?) and (?) is given by

\begin{split} &F=f_{1}^{l,m},\quad G=f_{2}^{l,m},\quad \overline{F}=f_{1}^{l,m+1},\quad \overline{G}=f_{2}^{l,m+1},\\ &t=q^{m-l}a_1,\quad c_1=qc^2,\quad c_2=q^la_0. \end{split}
Remark

Remark 3.2

WVK6

Each of equations (?) and (?) is equivalent to the q-P_{\rm V} (?). The correspondence between Equations (?) and (?) is given by

\begin{split} &F=f_{l,m}^{(1)},\quad G=f_{l,m}^{(3)},\quad \overline{F}=f_{l+1,m}^{(1)},\quad \overline{G}=f_{l+1,m}^{(3)},\\ &t=q^{l-m} {a_1}^{-1}{a_2}^{-1},\quad c_1=q^{-1}a_0a_1,\quad c_2={a_2}^{-1}{a_3}^{-1},\quad c_3=q^ma_0a_1{a_3}^{-1}, \end{split}

while that between Equations (?) and (?) is given by

\begin{split} &F=f_{l,m}^{(3)},\quad G=f_{l,m}^{(5)},\quad \overline{F}=f_{l,m+1}^{(3)},\quad \overline{G}=f_{l,m+1}^{(5)},\\ &t=q^{m} {a_3}^{-1}{a_4}^{-1},\quad c_1=q^{-1}a_2a_3,\quad c_2=q^{-l}{a_4}^{-1}{a_0}^{-1},\quad c_3=q^{-l}a_2a_3{a_0}^{-1}. \end{split}
Remark

Remark 3.3

BINM

Each of equations (?) and (?) is equivalent to the q-P_{\rm VI} (?). The correspondence between Equations (?) and (?) is given by

\begin{split} &F=\dfrac{a_4}{q^{2l}a_5}f_1^{l,m},\quad G=\dfrac{a_4}{q^{2l-1}a_5}g_1^{l,m},\quad \overline{F}=\dfrac{a_4}{q^{2l+2}a_5}f_1^{l+1,m},\quad \overline{G}=\dfrac{a_4}{q^{2l+1}a_5}g_1^{l+1,m},\\ &t=\dfrac{q^l{a_5}^{1/2}}{{a_4}^{1/2}},\quad c_1=(a_3a_4a_5)^{2},\quad c_2={a_3}^{2},\quad c_3=q^{2m}{a_1}^{2},\quad c_4=q^{-2m}{a_0}^{2}, \end{split}

while that between Equations (?) and (?) is given by

\begin{split} &F=\dfrac{a_0}{q^{2m}a_1}f_0^{l,m},\quad G=\dfrac{a_0}{q^{2m-1}a_1}g_0^{l,m},\quad \overline{F}=\dfrac{a_0}{q^{2m+2}a_1}f_0^{l,m+1},\quad \overline{G}=\dfrac{a_0}{q^{2m+1}a_1}g_0^{l,m+1},\\ &t=\dfrac{q^m{a_1}^{1/2}}{{a_0}^{1/2}},\quad c_1=(a_0a_1a_2)^{2},\quad c_2={a_2}^{2},\quad c_3=q^{-2l}{a_4}^{2},\quad c_4=q^{2l}{a_5}^{2}. \end{split}
Remark

Remark A.1

D97Q

Equations (?)–(?) are known as a q-discrete analogue of the Painlevé II equation[RG1996:MR1399286], that of the Painlevé III equation of type D_7^{(1)}[RGTT2000], that of the Painlevé III equation [CNP1991:MR1111648], that of the Painlevé V equation [SakaiH2001:MR1882403] and that of the Painlevé VI equation [JS1996:MR1403067], respectively.

Source and licence

The text on this page is from arXiv:2104.11433v2 by Nobutaka Nakazono, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.

The record of each item is in analysis/objects/2104/2104.11433.jsonl.