Hamiltonian structure, Darboux transformation for a soliton hierarchy associated with Lie algebra so(4, )
Wang Xin-Zeng†a),b), Dong Huan-Heb)
State Key Laboratory of Mining Disaster Prevention and Control Co-founded by Shandong Province and the Ministry of Science and Technology, Shandong University of Science and Technology (SDUST), Qingdao 266590, China
College of Mathematics and Systems Science, Shandong University of Science and Technology, Qingdao 266510, China

Corresponding author. E-mail: wangelxz@126.com

*Project supported by the National Natural Science Foundation of China (Grant Nos. 61170183 and 11271007), SDUST Research Fund, China (Grant No. 2014TDJH102), the Fund from the Joint Innovative Center for Safe and Effective Mining Technology and Equipment of Coal Resources, Shandong Province, the Promotive Research Fund for Young and Middle-aged Scientisits of Shandong Province, China (Grant No. BS2013DX012), and the Postdoctoral Fund of China (Grant No. 2014M551934).

Abstract

In this paper, we first introduce a Lie algebra of the special orthogonal group, g = so(4,), whose elements are 4 × 4 trace-free, skew-symmetric complex matrices. As its application, we obtain a new soliton hierarchy which is reduced to AKNS hierarchy and present its bi-Hamiltonian structure and Liouville integrability. Furthermore, for one of the equations in the resulting hierarchy, we construct a Darboux matrix T depending on the spectral parameter λ.

PACS: 02.20.Sv; 03.65.Aa
Keyword: zero curvature equation; recursion operator; Hamiltonian structure; Darboux transformation
1. Introduction

With the development of soliton theory, Lie algebra shows great promise. As it is well known, professor Tu presented a method which used Lie algebra as a tool to construct systematically the Lax equation. Especially, Tu[1] obtained trace identity-based loop algebra A1. Later, many kinds of Lie algebras and their subalgebras were constructed, and some hierarchies with physics meanings are given.[211] Recently, professor Ma[12] generated a hierarchy of soliton equations associated with the real Lie algebra so(3, ℝ ) and shows that each equation has a bi-Hamiltonian structure and integrable in the Liouville sense. The real Lie algebra of the so(3, ℝ ) is not isomorphic to the real Lie algebra sl(2, ℝ ) over the real field ℝ , and the soliton hierarchy[12] is not gauge-equivalent to the famous AKNS soliton hierarchy over the real field ℝ . In fact, the two Lie algebras are isomorphic to each other over the complex field , thus hierarchy[12] is equivalent to the AKNS hierarchy over the complex field .

It is known that Darboux transformation provides us with a purely algebraic and powerful method to find a new solution of a nonlinear equation from a known solution in Refs.  [13]– [16]. The classical Darboux transformation[17] mainly deals with a Sturm– Liouville problem, which presented solutions to the Korteweg-de Vries hierarchy of (1 + 1)-dimensional integrable systems with new potentials and new wave functions. Based on the work, proper transformation that applies to (2 + 1)-dimensional integrable systems was found. The method of the Darboux transformation has also been applied to the discrete integrable systems successfully. The present paper aims to review application of the Darboux transformation technique for superintegrable systems. In general, we assume every element of T is polynomial with respect to λ .

In this paper, we first introduce a Lie algebra of the special orthogonal group, g = so (4, ), which is a 4× 4 trace-free, skew-symmetric complex matrix set. As its application, we obtain a soliton hierarchy which is reduced to AKNS hierarchy and presented its Hamiltonian structures and Liouville integrability. Furthermore, for one of the equations in the resulting hierarchy, we present the Darboux matrix T, which the element is related to the spectral parameter λ according to the different position and number of spectral parameter λ .

2. Matrix loop algebra and soliton equations
2.1. Matrix loop algebra (4, ) and new soliton hierarchy

Firstly, let us consider the simple Lie algebra of the special orthogonal group, G = so(4, ), the Lie algebra of 4× 4 trace-free, skew-symmetric complex matrices. It has a basis

and the structure equations of so(4, ) are

We point out that the subalgebra G1 = span{e1, e2, e3} of so(4, ) isomorphic to so(3, ). We adopt the following matrix loop algebra

In the following, we work out a new soliton hierarchy from the matrix loop algebra (4, ), and introduce a spectral problem

Setting

then the stationary zero curvature equation

gives rise to a solution for V as follows:

Substituting

into Eqs.  (55), we obtain

On taking the initial values

and determining the sequence of {Ai,   Bi,   Ci,   Di,   Fi,   Gi, | i > 0}. Therefore, the first few sets can be computed as follows:

Now, taking

and reading a hierarchy of soliton equations:

from the zero curvature equations, we have

2.2. Hamiltonian structures and Liouville integrability for Eq.  (9)

It is direct to compute that

and read

Now the corresponding trace identity becomes

Balancing coefficients of each power of λ in the above equality we obtain

The case of m = 1 tells γ = 0, and thus we have

Consequently, we obtain the following Hamiltonian structure for the soliton hierarchy  (9)

with the Hamiltonian operator

and the Hamiltonian functionals

The recursion operator ϕ can be determined by the recursion relations  (6a)– (6f)

Equation  (9) has bi-Hamiltonian structure

with the Hamiltonian operator

2.3. The reductions of Eq.  (14)

Case 1r = s = 0

Case 2p = q = 0

where

and Hn, Hn-1 are the same as in Eq.  (13). From the above derivations, we know that the two cases have bi-Hamiltonian structure and are gauge-equivalent to the AKNS hierarchy over the complex field , respectively.

2.4. A nonlinear example

The first nonlinear integrable system in the hierarchy  (9) reads

It possesses the following bi-Hamiltonian structure

where the Hamiltonian pair {J, M} is defined by Eqs.  (12) and (15), and the Hamiltonian functionals, {H1, H2}, are given by

3. Darboux transformations of the spectral problem  (18)

We consider the spectral problem  (4) by introducing the temporal part

where

The compatibility condition ϕ xt = ϕ tx yields a zero curvature equation

Now we consider a Darboux transformation

where T is defined by

and

A new spectral problem reads

where U and N has the same form as U and N, except replacing p, q, r, and s with p, q, r, and s.

Now we consider the basic forms of T. First of all, we assume

where aij and bij (i, j = 1, 2, 3, 4) are functions of x and t. By inserting Eq.  (27) into Eq.  (24), we read that the coefficients matrix of λ 2 are

In the following, we consider the simplest non-trivial forms of T

3.1. Spatial scales of the Darboux transformation of Eq.  (18)

Under the condition equation  (29), we insert Eq.  (27) into Eq.  (24) again, compare the coefficients of λ j (j = 2, 1, 0), and know it is obviously trivial for the case of j = 2. We then have

for the coefficients matrix of λ (j = 1)

The case of j = 0 leads to the following equations

On the other hand, the solutions Φ , Φ of Eqs.  (4) and (26) are two 4× 4 matrices, thus det(Φ )= det(Φ )= const. from Tr(M)= Tr(M)= 0, which means that there is a constant λ = λ 1 and a solution ϕ = (ϕ 1, ϕ 2, ϕ 3, ϕ 4)T of Eq.  (4) which satisfy

By substituting Eqs.  (32), (42), (34), (40), (37), and (47) into Eq.  (30), we get

Combining Eq.  (30) with Eqs.  (48)– (52), we get

The unknown functions p, q can be defined by Eqs.  (36) and (38):

and unknown functions r, s can be defined by Eqs.  (44) and (46)

The others of Eqs.  (32)– (47) can be proved to be satisfied automatically (under the condition ϕ x = ).

3.2. Temporal scales of the Darboux transformation

If one can prove that the transformation equation  (27) maps Eq.  (21) into ϕ t = Nϕ , where N has the same form as N in Eq.  (21) except changing p, q, r, and s with p, q, r, and s, then the compatibility condition ϕ xt = ϕ tx holds, i.e.,

so that (p, q, r, s) is a new solution of Eq.  (21).

The main idea of the proof is following: here we only need to prove equation ϕ t = Nϕ .

Let us compare the coefficient of λ j (j = 3, 2, 1, 0). It is trivial for the cases of j = 3. For the case of j = 2, the coefficients matrix is

We have the same form as Eq.  (31) when a22x= 0.

The case of j = 1 leads to the following equations

After substituting Eqs.  (32)– (38), (40)– (44), (46), (47) into Eqs.  (59)– (65), (67)– (71), (73), (74) respectively, and replacing Eqs.  (66) and (72) with Eq.  (30), then we find equations  (59)– (74) are valid.

Also the following equations are presented from the case of j = 0

Here, we only give the proof of Eq.  (88) for brevity.

From Eq.  (53) and ϕ x= , we read

Also, from Eqs.  (56) and (57), we have

Again, from Eqs.  (53) and ϕ t = , we propose that

We know that equation  (88) is true by substituting Eqs.  (53)– (57), (92)– (94) into Eq.  (88).

4. Remark

As the subalgebras of so(4, ), G1 = span{e1, e2, e3} and G2 = span{e1, e5, e6}, both are isomorphic to so(3, ), and both are isomorphic to sl(2, ) over the complex field , we know that equations  (16) and (17), as the reductions of Eq.  (11), are equivalent to AKNS hierarchy over the complex field . Therefore, equation  (9) was regarded as the integrable coupling of AKNS and itself, which was also called the bi-AKNS system.

In the past, general coupling form[1823] by making use of Lie algebra A1, A2 and their subalgebras was presented in the following

and general super-integrable hierarchy form through use of Lie subalgebra B(0, 1) was proposed by Li, [24] Ma[25] as follows:

In this paper, we obtained the integrable coupling by making use of orthogonal Lie algebra so(4, ), which possessed the form of Eq.  (96), and is not super-integrable coupling. It is the first time that we have found the results.

Acknowledgments

Authors would like to thank Professor Ma Wen-Xiu for his useful discussion during his visit to Shandong University of Science and Technology of China.

Reference
1 Tu G Z 1989 J. Math. Phys. 30 330 DOI:10.1063/1.528449 [Cited within:1]
2 Ma W X, Fuchssteiner B and Oevel W 1996 Phys. Lett. A 233 331 DOI:10.1016/S0378-4371(96)00225-7 [Cited within:1]
3 Fan E G 2002 Acta Math. Appl. Sin. 18 405 [Cited within:1]
4 Meng J H and Ma W X 2013 Commun. Theor. Phys. 59 385 DOI:10.1088/0253-6102/59/4/01 [Cited within:1]
5 Xia T C, You F C and Chen D Y 2005 Chaos, Solitons and Fractals 23 1911 DOI:10.1016/j.chaos.2004.07.004 [Cited within:1]
6 Ma W X 2013 British J. Appl. Sci. Tech. 3 1336 DOI:10.9734/BJAST [Cited within:1]
7 Zhang Y F and Tam H W 2008 Commun. Nonlinear Sci. Numer. Simul. 13 524 DOI:10.1016/j.cnsns.2006.06.003 [Cited within:1]
8 Ma W X, Xu X X and Zhang Y F 2006 Phys. Lett. A 351 125 DOI:10.1016/j.physleta.2005.09.087 [Cited within:1]
9 Xia T C and You F C 2007 Chin. Phys. 16 605 DOI:10.1088/1009-1963/16/3/008 [Cited within:1]
10 Yang H W, Dong H H and Yin B S 2012 Chin. Phys. B 21 100204 DOI:10.1088/1674-1056/21/10/100204 [Cited within:1]
11 Zhang Y F and Liu J 2008 Commun. Theor. Phys. 50 289 DOI:10.1088/0253-6102/50/2/01 [Cited within:1]
12 Ma W X 2013 Appl. Math. Comput. 220 117 DOI:10.1016/j.amc.2013.04.062 [Cited within:3]
13 Li Y S, Ma W X and Zhang J E 2000 Phys. Lett. A 275 60 DOI:10.1016/S0375-9601(00)00583-1 [Cited within:1]
14 Zeng Y B and Li Y S 1996 Acta Math. Sin. 12 217 DOI:10.1007/BF02108164 [Cited within:1]
15 Zhang Y F, Han Z and Tam H W 2013 Appl. Math. Comput. 219 5837 DOI:10.1016/j.amc.2012.11.086 [Cited within:1]
16 Fan E G and Chow K W 2011 J. Math. Phys. 52 023504 DOI:10.1063/1.3545804 [Cited within:1]
17 Matveev V B and Salle M A 1991 Darboux Transformations and Solitons Berlin Springer-Verlag 108 121 [Cited within:1]
18 Guo F K and Zhang Y F 2005 J. Phys. A: Math Gen. 38 8537 DOI:10.1088/0305-4470/38/40/005 [Cited within:1]
19 Ma W X 2012 Chin. Ann. Math. B 33 207 DOI:10.1007/s11401-012-0702-7 [Cited within:1]
20 Dong H H and Wei Y 2007 J. Math. Phys. 48 053501 DOI:10.1063/1.2719140 [Cited within:1]
21 Zhang Y F and Yan Q Y 2003 Acta Phys. Sin. 52 2109 (in Chinese) [Cited within:1]
22 Ma W X 2011 Appl. Math. Comput. 217 7238 DOI:10.1016/j.amc.2011.02.014 [Cited within:1]
23 Zhang Y F 2003 Chin. Phys. 12 1194 DOI:10.1088/1009-1963/12/11/302 [Cited within:1]
24 Li Y S and Zhang L N 1990 J. Math. Phys. 31 470 DOI:10.1063/1.528881 [Cited within:1]
25 Ma W X, He J S and Qin Z Y 2008 J. Math. Phys. 49 033511 DOI:10.1063/1.2897036 [Cited within:1]