Applied Mathematics and Computation 216 (2010) 1965–1971

Contents lists available at ScienceDirect

Applied Mathematics and Computation

journal homepage: www.elsevier.com/locate/amc

 0
The improved GG -expansion method and its applications to the
Broer–Kaup equations and approximate long water wave equations
Shimin Guo, Yubin Zhou, Chenxia Zhao *
School of Mathematics and Statistics, Lan-zhou University, Lan-zhou 730000, China

a r t i c l e i n f o a b s t r a c t
 0
Keywords:   By introducing a new general ansätze, the improved GG -expansion method is proposed to
0
Improved GG -expansion method construct exact solutions of both Broer–Kaup equations and approximate long water wave
Solitary wave solution equations. As a result, some new travelling wave solutions involving parameters, expressed
Nonlinear evolution equations by three types of functions which are the hyperbolic functions, the trigonometric functions
Broer–Kaup equations
and the rational functions, are obtained. When the parameters are taken as special values,
Approximate long water wave equations
the solitary wave solutions are derived from the hyperbolic function solutions. The pro-
posed method is straightforward, concise and effective, and can be applied to other nonlin-
ear evolution equations in mathematical physics.
Ó 2010 Elsevier Inc. All rights reserved.

1. Introduction

In the present paper, we will seek exact solutions of the following two nonlinear evolution equations (NLEEs): one is the
Broer–Kaup (BK) equations [1,2] in the form

ut þ uux þ v x ¼ 0;
ð1:1Þ
v t þ ux þ ðuv Þx þ uxxx ¼ 0;
which is used to model the bi-directional propagation of long waves in shallow water, and another is the approximate long
water wave (ALWW) equations [2–4] in the form

ut  uux  v x þ auxx ¼ 0;
ð1:2Þ
v t  ðuv Þx  av xx ¼ 0:
Because of the importance of Eqs. (1.1) and (1.2) in the field of mathematical physics, it is a significant task to search for
explicit solutions of the above two equations. Up to now, many powerful methods have been established and developed by
mathematicians and physicists to find exact solutions of NLEEs, such as the inverse scattering method [5], the truncated
Painlevé expansion [6], the Bäcklund transformation [7], the Darboux transformation [8], the Lie symmetries method [9],
the homogenous balance method [10–12], and so on.
In recent years, with the development of symbolic computation packages like Maple and Mathematica, which enable us
to perform the tedious and complex computation on computer, much work has been focused on the direct methods [13–24]
 0
to construct exact solutions of NLEEs. The GG -expansion method proposed by Wang et al. [25], is one of the most effective
direct methods to obtain travelling wave solutions of a large number of NLEEs, such as the KdV equation, the mKdV equation,
the variant Boussinesq equations, the Hirota–Satsuma equations, and so on. Later, the further developed methods named the

* Corresponding author.
E-mail address: mybzhou@yahoo.com.cn (C. Zhao).

0096-3003/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.amc.2010.03.026
1966 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971

 0  0  0
generalized GG -expansion method, the modified GG -expansion method and the extended GG -expansion method have been
proposed in Refs. [26–28], respectively. As we know, when using the direct method, the choice of an appropriate ansätz is of
0
great importance . In this paper, by introducing a new general ansätze, we propose the improved GG -expansion method,
which can be used to obtain travelling wave solutions of NLEEs.  0
The rest of this paper is organized as follows: in the following Section 2, the improved GG -expansion method for finding
out exact solutions of NLEEs is described. In Section 3, we illustrate the applications of this method to the BK equation (1.1)
and ALWW equation (1.2). In Section 4, some conclusions are given.

G 0 
2. Description of the improved G
-expansion method

In this section, we give the detailed description of our method. Suppose that a nonlinear evolution equation, say in two
independent variables x and t, is given by

Pðu; ut ; ux ; utt ; uxx ; uxt ; . . .Þ ¼ 0; ð2:1Þ

where u ¼ uðx; tÞ is an unknown function, P is a polynomial in u ¼ uðx; tÞ and its various partial derivatives, in which the
highest order derivatives and nonlinear terms are involved. To determine u explicitly, we take the following five steps:

Step 1: Use the travelling wave transformation:

uðx; tÞ ¼ uðnÞ; n ¼ x  Vt; ð2:2Þ

where V is a constant to be determined latter. Then, the NLEE (2.1) is reduced to a nonlinear ordinary differential
equation (NLODE) for u ¼ uðnÞ:

Pðu; Vu0 ; u0 ; V 2 u00 ; u00 ; Vu00 ; . . .Þ ¼ 0: ð2:3Þ

Step 2: We suppose that the NLODE (2.3) has the following solution:
G0 i G0 i
X
1
ai X
n
ai
G G
uðnÞ ¼  G0 i þ a0 þ   0 i ; ð2:4Þ
i¼n 1þr G i¼1 1 þ r GG
where r; ai ði ¼ n; n þ 1; . . . ; n  1; nÞ are constants to be determined later, n is a positive integer, and G ¼ GðnÞ
satisfies the following second order linear ordinary differential equation(LODE):

G00 þ lG ¼ 0; ð2:5Þ
where l is a real constant. The general solutions of Eq. (2.5) can be listed as follows.When l < 0, we obtain the
hyperbolic function solution of Eq. (2.5)
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
GðnÞ ¼ A1 coshð lnÞ þ A2 sinhð lnÞ; ð2:6Þ
where A1 and A2 are arbitrary constants.When l > 0, we obtain the trigonometric function solution of Eq. (2.5)
pffiffiffiffi pffiffiffiffi
GðnÞ ¼ A1 sinð lnÞ þ A2 cosð lnÞ; ð2:7Þ
where A1 and A2 are arbitrary constants.When l ¼ 0, we obtain the rational function solution of Eq. (2.5)
GðnÞ ¼ A1 þ A2 n; ð2:8Þ
where A1 and A2 are arbitrary constants.
Step 3: Determine the positive integer n by balancing the highest order derivatives and nonlinear terms in Eqs. (2.1) or
(2.3) (see Refs. [10–12,25] for details).  0 k
Step 4: Substituting (2.4) along with Eq. (2.5) into Eq. (2.3) and then setting all the coefficients of GG ðk ¼ 1; 2; . . .Þ of the
resulting system’s numerator to zero, yields a set of over-determined nonlinear algebraic equations for V; r and
ai ði ¼ n; n þ 1; . . . ; n  1; nÞ.
Step 5: Assuming that the constants V; r; ai ði ¼ n; n þ 1; . . . ; n  1; nÞ can be obtained by solving the algebraic equa-
tions in Step 4, then substituting these constants and the known general solutions of Eq. (2.5) into (2.4), we
can obtain the explicit solutions of Eq. (2.1) immediately.

G0 
3. Applications of the improved G
-expansion method

In this section, we will construct travelling wave solutions of the Broer–Kaup equation (1.1) and the approximate long
water wave Eq. (1.2) by using the method described in Section 2.
 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971 1967

3.1. The explicit solutions of the Broer–Kaup equations

By considering the wave transformations uðx; tÞ ¼ uðnÞ; v ðx; tÞ ¼ v ðnÞ and n ¼ x  Vt, where V is a constant to be deter-
mined latter, we change the BK equation (1.1) to the following NLODEs
(
Vu0 þ uu0 þ v 0 ¼ 0;
ð3:1Þ
V v 0 þ u0 þ ðuv Þ0 þ u000 ¼ 0:

Integrating the NLODEs above with respect to n once yields

(
C 1  Vu þ 12 u2 þ v ¼ 0;
ð3:2Þ
C 2  V v þ u þ uv þ u00 ¼ 0;

where C 1 and C 2 are integration constants to be determined latter.

By balancing the highest order derivative terms and nonlinear terms in Eq. (3.2), we suppose that Eq. (3.2) own the solu-
tions in the form
8 0 0
a1 ðG Þ b1 ð1þrðGG ÞÞ
>
> uðnÞ ¼ a0 þ 1þr GG0 þ ;
>
< ð Þ G0
ð Þ
G G
ð3:3Þ
>
> ð Þ
c1 GG
0 0
d1 ð1þrðGG ÞÞ
0 2
c2 ðGG Þ
0
d2 ð1þrðGG ÞÞ
2
>
: v ðnÞ ¼ c0 þ þ
0 0 þ 2 þ ;
rð Þ 0 2
1þ GG ðGG Þ 0
ð1þrð G ÞÞ
G
ðGG Þ

where G ¼ GðnÞ satisfies Eq. (2.5), r; a0 ; a1 ; b1 ; c0 ; c1 ; c2 ; d1 and d2 are constants to be determined

 0 klatter.
Substituting (3.3) along with Eq. (2.5) into Eq. (3.2) and then setting all the coefficients of GG ðk ¼ 1; 2; . . .Þ of the result-
ing system’s numerator to zero, yields a set of over-determined nonlinear algebraic equations about r; a0 ; a1 ; b1 ;
c0 ; c1 ; c2 ; d1 ; d2 ; V; C 1 and C 2 . Solving the over-determined algebraic equations by Maple or Mathematica, we can obtain
the following results:

Case 1:
1
C 1 ¼ 2l2 r2 þ a20  2a0 lr þ 1 þ 2l; C 2 ¼ a0  2lr; V ¼ a0  2lr; a1 ¼ 2ð1 þ lr2 Þ;
2
b1 ¼ 0; c0 ¼ 1  2l2 r2  2l; c1 ¼ 4lrð1 þ lr2 Þ; c2 ¼ 2ð1 þ lr2 Þ2 ; d1 ¼ d2 ¼ 0;

where a0 ; r are arbitrary constants.

Case 2:
1
C 1 ¼ 1 þ a20 þ 8l; C 2 ¼ a0 ; V ¼ a0 ; a1 ¼ 2; b1 ¼ 2l; c0 ¼ 1  4l;
2
c1 ¼ 0; c2 ¼ 2; d1 ¼ 0; d2 ¼ 2l2 ; r ¼ 0;
where a0 is an arbitrary constants.
Case 3:
1
C 1 ¼ 2l2 r2 þ a20  2a0 lr þ 1 þ 2l; C 2 ¼ a0  2lr; V ¼ a0  2lr; a1 ¼ 0; b1 ¼ 2l;
2
c0 ¼ 1  2l2 r2  2l; c1 ¼ 0; c2 ¼ 0; d1 ¼ 4l2 r; d2 ¼ 2l2 ;
where a0 ; r are arbitrary constants.
Case 4:
1
C 1 ¼ 4l2 r2 þ a20  2a0 lr þ 1  4l; C 2 ¼ 8l3 r3  8l2 r  a0  2lr; V ¼ a0  2lr;
2
a1 ¼ 2ð1 þ lr2 Þ; b1 ¼ 2l; c0 ¼ 1; c1 ¼ 4lrð1 þ lr2 Þ; c2 ¼ 2ð1 þ lr2 Þ2 ;
d1 ¼ 4l2 r; d2 ¼ 2l2 ;
where a0 ; r are arbitrary constants.

Using Case 4, (3.3) and the general solutions of Eq. (2.5), we can find the following travelling wave solutions of BK equa-
tion (1.1).
When l < 0, we obtain the hyperbolic function solutions of Eq. (1.1)
1968 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971

8 pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi

> ð1þlr2 Þ lðA1 sinh lnþA2 cosh lnÞ
>
> uðnÞ ¼ a0  2 ðA1 þA2 rpffiffiffiffiffi  lÞ cosh
pffiffiffiffiffi
 l nþðA þA r
pffiffiffiffiffi
 l Þ
pffiffiffiffiffi
sinh ln
>
> 2 1
>
> p ffiffiffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
>
> ðA þA r lÞ cosh lnþðA þA r lÞ sinh ln
2 l 1 2 A1 sinh pffiffiffiffiffilnþA2 2cosh1 pffiffiffiffiffiln ;
>
>
>
>
< pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
4lrð1þlr2 Þ lðA1 sinh lnþA2 cosh lnÞðA1 cosh lnþA2 sinh lnÞ
v ðnÞ ¼ 1 þ pffiffiffiffiffi pffiffiffiffiffi
ððA1 þA2 r lÞ cosh lnþðA2 þA1 r lÞ sinh lnÞ2
pffiffiffiffiffi pffiffiffiffiffi ð3:4Þ
>
>
>
> 2 r4 1ÞðA sinh pffiffiffiffiffi
lnþA2 cosh lnÞ2
pffiffiffiffiffi
>
>  ððA þA2lrðlpffiffiffiffiffi 1
pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
>
> lÞ cosh lnþðA2 þA1 r lÞ sinh lnÞ2
>
> 1 2
>
> pffiffiffiffiffi pffiffiffiffiffi 2 pffiffiffiffiffi pffiffiffiffiffi
>
: 2l r ðA1 sinh lnþA2 cosh lnÞ þ2lðA1 cosh lnþA2 sinh lnÞ2
2 2
þ pffiffiffiffiffi
ðA sinh lnþA cosh lnÞ2
pffiffiffiffiffi ;
1 2

where n ¼ x  Vt; V ¼ a0  2lr; a0 ; r; A1 ; A2 are arbitrary constants.

In particular, setting r ¼ A1 ¼ 0; A2 –0, the following solitary wave solutions of BK equation (1.1) are discovered
( pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
uðnÞ ¼ a0  2 l coth ln  2 l tanh ln;
ð3:5Þ
v ðnÞ ¼ 1 þ 2lcoth2 pffiffiffiffiffiffiffi 2 pffiffiffiffiffiffiffi
ln þ 2ltanh ln;

where n ¼ x  a0 t; a0 is an arbitrary constants.

Setting again r ¼ 0; A1 > 0; A21 > A22 , then uðnÞ and v ðnÞ become the solitary wave solutions of Eq. (1.1) as follows
( pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
uðnÞ ¼ a0  2 l tanhð ln þ n0 Þ  2 l cothð ln þ n0 Þ;
ð3:6Þ
v ðnÞ ¼ 1 þ 2ltanh2 ðpffiffiffiffiffiffiffi 2 pffiffiffiffiffiffiffi
ln þ n0 Þ þ 2lcoth ð ln þ n0 Þ;
1
where n ¼ x  a0 t; a0 is an arbitrary constants, n0 ¼ tanh AA21 . It is easy to see that if A1 ; A2 ; l; r are taken as other special val-
ues in a proper way, more solitary wave solutions of Eq. (1.1) can be obtained, here we omit them for simplicity.
When l > 0, we get the trigonometric function solutions of Eq. (1.1)
8 pffiffiffi pffiffiffi pffiffiffi
> ð1þlr2 Þ lðA cos lnA sin lnÞ
>
> uðnÞ ¼ a0  2 ðA1 A2 rpffiffilffiÞ sin p1ffiffilffinþðA2 þA1 r2 pffiffilffiÞ cos pffiffilffin
>
>
>
> pffiffiffiffi pffiffiffi pffiffiffi pffiffiffi pffiffiffi
>
> 2 l ðA1 A2 r Al1Þ sin lnþðA2 þA1 r lÞ cos ln
pffiffiffi pffiffiffi ;
>
> cos lnA2 sin ln
>
>
< 2 pffiffiffiffiffi pffiffiffi pffiffiffi pffiffiffi pffiffiffi
v ðnÞ ¼ 1 þ 4lrð1þlrððAÞ1 A lðA1 cos lnA2 sin lnÞðA1 sin lnþA2 cos lnÞ
pffiffiffi pffiffiffi pffiffiffi
2 r lÞ sin lnþðA2 þA1 r lÞ cos lnÞ
pffiffiffi 2 ð3:7Þ
>
>
>
> pffiffiffi
2lðl2 r4 1ÞðA cos lnA sin lnÞ2
pffiffiffi
>
> þ ððA A rpffiffilffiÞ sin p1ffiffilffinþðA þA r2pffiffilffiÞ cos pffiffilffinÞ2
>
>
>
> 1 2 2 1
>
> pffiffiffi pffiffiffi 2 pffiffiffi pffiffiffi 2
> 2 2
: þ 2l r ðA1 cos lnA 2 sin lnÞ 2lðA1 sin lnþA2 cos lnÞ
pffiffiffi
ðA cos lnA sin lnÞ2
pffiffiffi ;
1 2

where n ¼ x  Vt; V ¼ a0  2lr; a0 ; r; A1 ; A2 are arbitrary constants.

When l ¼ 0, we get the rational function solutions of Eq. (1.1)
8 2A
< uðnÞ ¼ a0  A1 þA2 r2þA2 n ;
ð3:8Þ
: v ðnÞ ¼ 1  2A22
ðA1 þA2 rþA2 nÞ 2 ;

where n ¼ x  a0 t; a0 ; r; A1 ; A2 are arbitrary constants.

Using Case 1–Case 3, (3.3) and the general solutions of Eq. (2.5), we could obtain more exact solutions of Eq. (1.1), and
here we do not list all of them.

3.2. The explicit solutions of the approximate long water wave equations

Now, we will consider the approximate long water wave equation (1.2). Making the transformation
uðx; tÞ ¼ uðnÞ; v ðx; tÞ ¼ v ðnÞ; n ¼ x  Vt, where V is a constant to be determined latter, we change the ALWW equation (1.2)
to the form

Vu0  uu0  v 0 þ au00 ¼ 0;
ð3:9Þ
V v 0  ðuv Þ0  av 00 ¼ 0:
Integrating the NLODEs above with respect to n once yields
(
C 1  Vu  12 u2  v þ au0 ¼ 0;
ð3:10Þ
C 2  V v  uv  av 0 ¼ 0;
where C 1 and C 2 are integration constants to be determined latter.
By balancing the highest order derivative terms and nonlinear terms in Eq. (3.10), we suppose that the solutions of Eq.
(3.10) is of the form
 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971 1969

8 0 0
a1 ðGG Þ b1 ð1þrðGG ÞÞ
>
< uðnÞ ¼ a0 þ 1þrðGG0 Þ þ
> G0
ðGÞ
;
0 0 0 2 0 2 ð3:11Þ
>
> c1 ðGG Þ d1 ð1þrðGG ÞÞ c2 ðGG Þ d2 ð1þrðGG ÞÞ
: v ðnÞ ¼ c0 þ 0 þ G0
þ 2 þ 2 ;
1þrðGG Þ ð Þ
G ð1þrð ÞÞ G0 G0
ð Þ
G G

where G ¼ GðnÞ satisfies Eq. (2.5), r; a0 ; a1 ; b1 ; c0 ; c1 ; c2 ; d1 ; d2 are constants to determined latter.  0 k

Substituting (3.11) along with Eq. (2.5) into Eq. (3.10) and then setting all the coefficients of GG ðk ¼ 1; 2; . . .Þ of the
resulting system’s numerator to zero, yields a set of over-determined nonlinear algebraic equations about
r; a0 ; a1 ; b1 ; c0 ; c1 ; c2 ; d1 ; d2 ; V; C 1 and C 2 . Solving the over-determined algebraic equations by Maple or Mathematica, we
can obtain the following results:

Case 1:
1
C 1 ¼ 2a2 l  2a2 r2 l2  a20  2alra0 ; C 2 ¼ 0; V ¼ 2alr  a0 ; a1 ¼ 2að1 þ r2 lÞ;
2
b1 ¼ 0; c0 ¼ 4a2 lð1 þ lr2 Þ; c1 ¼ 8ra2 lð1 þ lr2 Þ; c2 ¼ 4a2 ð1 þ lr2 Þ2 ; d1 ¼ d2 ¼ 0;
where a0 ; r are arbitrary constants.
Case 2:
1
C 1 ¼ 8a2 l  a20 ; C 2 ¼ 0; V ¼ a0 ; a1 ¼ 2a; b1 ¼ 2al; c0 ¼ 8a2 l;
2
c1 ¼ 0; c2 ¼ 4a2 ; d1 ¼ 0; d2 ¼ 4a2 l2 ; r ¼ 0
where a0 is an arbitrary constant.
Case 3:
1
C 1 ¼ 2a2 l  2a2 r2 l2  a20 þ 2alra0 C 2 ¼ 0; V ¼ 2alr  a0 ; a1 ¼ 0; b1 ¼ 2al;
2
c0 ¼ 4a2 lð1 þ lr2 Þ; c1 ¼ c2 ¼ 0; d1 ¼ 8a2 l2 r; d2 ¼ 4a2 l2 ;
where a0 ; r are arbitrary constants.
Case 4:
1
C 1 ¼ 4a2 l þ 4a2 r2 l2  a20 þ 2alra0 ; C 2 ¼ 8a3 l2 rð1 þ lr2 Þ; V ¼ 2alr  a0 ;
2
a1 ¼ 2að1 þ r2 lÞ; b1 ¼ 2al; c0 ¼ c1 ¼ c2 ¼ 0; d1 ¼ 8a2 l2 r; d2 ¼ 4a2 l2 ;
where a0 ; r are arbitrary constants.
Case 5:
1
C 1 ¼ 4a2 l þ 4a2 r2 l2  a20  2alra0 ; C 2 ¼ 8a3 l2 rð1 þ lr2 Þ; V ¼ 2alr  a0 ;
2
a1 ¼ 2að1 þ lr2 Þ; b1 ¼ 2al; c0 ¼ 0; c1 ¼ 8ra2 lð1 þ lr2 Þ; c2 ¼ 4a2 ð1 þ lr2 Þ2 ;
d1 ¼ d2 ¼ 0;
where a0 ; r are arbitrary constants.

Using Case 5, Eq. (3.11) and the general solutions of Eq. (2.5), the following exact solutions of ALWW Eq. (1.2) are
obtained.
When l < 0, we obtain the hyperbolic function solutions of Eq. (1.2)
8 pffiffiffiffiffi pffiffiffiffiffi
2að1þlr2 Þ lðA1 sinh lnþA2 cosh lnÞ
pffiffiffiffiffi
>
> uðnÞ ¼ a0 þ ðA1 þA2 rpffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
>
> lÞ cosh lnþðA2 þA1 r lÞ sinh ln
pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
>
> p ffiffiffiffiffiffiffi r l l r lÞ sinh ln
>
< 2a l ðA 1 þA 2  Þ cosh 
pffiffiffiffiffi
nþðA 2 þA 1 
pffiffiffiffiffi ;
A1 sinh lnþA2 cosh ln
r 2 lð1þlr2 Þp ffiffiffiffiffilðA sinh pffiffiffiffiffilnþA cosh pffiffiffiffiffilnÞðA cosh pffiffiffiffiffi l
pffiffiffiffiffi ð3:12Þ
> 8 a  nþA 2 sinh lnÞ
>
> v ðnÞ ¼ 1
pffiffiffiffiffi pffiffiffiffiffi
2
ððA1 þA2 r lÞ cosh lnþðA2 þA1 r lÞ sinh lnÞ2
1
pffiffiffiffiffi pffiffiffiffiffi
>
>
>
> pffiffiffiffiffi
4lr2 ðl2 r4 1ÞðA1 sinh lnþA2 cosh lnÞ2
pffiffiffiffiffi
:  ððA þA pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi ;
1r lÞ cosh lnþðA þA r lÞ sinh lnÞ2
2 2 1

where n ¼ x  Vt; V ¼ 2alr  a0 ; a0 ; r; A1 ; A2 are arbitrary constants.

In particular, when setting r ¼ A1 ¼ 0; A2 –0, the solutions (3.12) can be written as
( pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
uðnÞ ¼ a0 þ 2a l coth ln  2a l tanh ln;
ð3:13Þ
v ðnÞ ¼ 4la2 coth2 pffiffiffiffiffiffiffi
ln;

which is the solitary wave solution of Eq. (1.2), where n ¼ x þ a0 t; a0 is an arbitrary constant.
1970 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971

Setting again r ¼ 0; A1 > 0; A21 > A22 , the following solitary wave solutions of ALWW equation (1.2) are obtained
( pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
uðnÞ ¼ a0 þ 2a l tanhð ln þ n0 Þ  2a l cothð ln þ n0 Þ;
ð3:14Þ
v ðnÞ ¼ 4la2 tanh2 ðpffiffiffiffiffiffiffi
ln þ n0 Þ;
1
where n ¼ x þ a0 t; a0 is an arbitrary constant, and n0 ¼ tanh AA21 . Clearly, when setting A1 ; A2 ; l; r to other special values in a
proper way, we can get more solitary wave solutions of ALWW Eq. (1.2), and here we do not list all of them.
When l > 0, we obtain the trigonometric function solutions of Eq. (1.2)
8 pffiffiffi pffiffiffi
að1þlr2 Þ lðA1 cos lnA2 sin lnÞ
pffiffiffi
>
> uðnÞ ¼ a0 þ ðA12A pffiffiffi pffiffiffi pffiffiffi pffiffiffi
>
> 2 r lÞ sin lnþðA2 þA1 r lÞ cos ln
pffiffiffi pffiffiffi pffiffiffi pffiffiffi
>
> p ffiffiffi
ffi r l sin l r l cos ln
>
<
ðA A Þ
þ2a l 1 2 A1 cos pffiffilffinA22sin 1pffiffilffin
nþðA þA Þ
;
2 2 pffiffiffi pffiffiffi pffiffiffi pffiffiffi pffiffiffi ð3:15Þ
>
>
> v ðnÞ ¼ 8ra lð1þlr Þ lðA1 cos lnA2 sin lnÞðA1 sin lnþA2 cos lnÞ
pffiffiffi pffiffiffi pffiffiffi
ððA1 A2 r lÞ sin lnþðA2 þA1 r lÞ cos lnÞ2
pffiffiffi
>
>
>
> pffiffiffi
4lr2 ðl2 r4 1ÞðA cos lnA sin lnÞ2
pffiffiffi
: þ ððA A rpffiffilffiÞ sin pffiffilffi1nþðA þA rp2 ffiffilffiÞ cos pffiffilffinÞ2 ;
1 2 2 1

where n ¼ x  Vt; V ¼ 2alr  a0 ; a0 ; r; A1 ; A2 are arbitrary constants.

When l ¼ 0, we get the rational function solutions of Eq. (1.2)
8
< uðnÞ ¼ a0 þ A1 þA2a2 rA2þA2 n ;
4a2 A22
ð3:16Þ
: v ðnÞ ¼  ;
ðA1 þA2 rþA2 nÞ2

where n ¼ x þ a0 t; a0 ; r; A1 ; A2 are arbitrary constants.

Using Case 1–Case 4 , (3.11) and the general solutions of Eq. (2.5), more travelling wave solutions of ALWW equation (1.2)
can be obtained. Here we omit them for simplicity.
 0
Remark 1. The ansätz (2.4) proposed in this paper is more general than the ansätz in GG -expansion method [25] and
 0
modified GG -expansion method [27]. If we set the parameters in (2.4) and (2.5) to special values, the above two methods
 0
can be recovered by our proposed method. Therefore, the new method is more powerful than the GG -expansion method
 0
and modified GG -expansion method, and some new types of travelling wave solutions and solitary wave solutions would be
expected for some NLEEs.
0
ðG Þ
Remark 2. By writing the exact solutions of NLEEs as polynomials of 1þrG G0 , the equations can be changed into the nonlinear
ðGÞ  0
system of algebraic equations, which can be solved with the help of symbolic computation. Therefore, the improved GG -
expansion method is a pure algebraic algorithm which can be applied to integrable system and non-integrable system.

Remark 3. We can further extend our method by introducing a more generalized ansätz as follows:
P  0 r 1  0 r 2
 0 rm P  0 s1  0 s2  0 sm
G1 G2 G1 G2
X
n    GGmm
r1 þr2 þþr m ¼k ar 1 r 2 r m G1 Xn
G2 s1 þs2 þþsm ¼l bs1 s2 sm G1 G2
   GGmm
uðnÞ ¼ a0 þ   0  0  0 k þ   0  0  0 l ; ð3:17Þ
G G G G
k¼1 1 þ r1 G11 þ r2 G22 þ    þ rm GGmm l¼1 1 þ r1 G11 þ r2 G22 þ    þ rm GGmm

where r i P 0; si 6 0ði ¼ 1; 2; . . . ; mÞ; n ¼ x  Vt; V; ar1 r2 rm ; bs1 s1 sm ; ri ði ¼ 1; 2;    ; mÞ are constants to be determined, the posi-
tive integer n can be determined by balancing the highest order derivatives and nonlinear terms in the given NLEEs,
Gk ¼ Gk ðnÞðk ¼ 1; 2; . . . ; mÞ satisfy the following second order ordinary partial differential equations:

G00k þ lk Gk ¼ 0; k ¼ 1; 2; . . . ; m ð3:18Þ
where lk ðk ¼ 1; 2; . . . ; mÞ are real constants, and li –lj ði–jÞ. It is clear that by using ansätz (3.17), we could get more novel
travelling wave solutions, complexiton solutions [29] and solitary wave solutions of NLEEs.

4. Conclusion
 0
In this paper, the improved GG -expansion method has been proposed to find out exact solutions of NLEEs based on a new
general ansätze. By using the proposed method, we have successfully obtained some travelling wave solutions of both Broer–
Kaup equations and approximate long water wave equations. These exact solutions include the hyperbolic function solu-
tions, trigonometric function solutions and rational function solutions. When the parameters are taken as special values,
the solitary wave solutions are derived from the hyperbolic function solutions. To the best of our knowledge, the solutions
0
obtained in this paper have not been reported in previous literature. The remarks illustrate that the extended GG -expansion
 0
method is more powerful than the methods in Refs. [25,27]. The work shows that the improved GG -expansion method is
 S. Guo et al. / Applied Mathematics and Computation 216 (2010) 1965–1971 1971

direct, concise and effective, and can be applied to other NLEEs in mathematical physics, such as the (2 + 1)-dimensional
Painlevé integrable Burgers equations [30], the (2 + 1)-dimensional breaking soliton equations [31,32], the (2 + 1)-dimen-
sional Boiti-Leon–Pempinelli equations [32], the (2 + 1)-dimensional Broer–Kaup–Kupershmidt (BKK) equations [33], the
(2 + 1)-dimensional Burgers equations [34,35], and so on.

References

[1] D.J. Kaup, A higher order water wave equation and method for solving it, Progress of Theoretical Physics 54 (1975) 396–408.
 0
[2] Mingliang Wang, Jinliang Zhang, Xiangzheng Li, Application of the GG -expansion to travelling wave solutions of the Broer–Kaup and the approximate
long water wave equations, Applied Mathematics and Computation 206 (2008) 321–326.
[3] G.B. Whitham, Variational methods and application to water waves, Proceedings of the Royal Society of London Series A 299 (1967) 6–25.
[4] L.J.F. Broer, Approximate equations for long water waves, Applied Scientific Research 31 (1975) 377–395.
[5] M.J. Ablowitz, P.A. Clarkson, Solitons, Nonlinear Evolution Equations and Inverse Scattering Transform, Cambridge University Press, Cambridge, 1991.
[6] F. Cariello, M. Tabor, Similarity reductions from extended Painlevé expansions for nonintegrable evolution equations, Physica D 53 (1991) 59–70.
[7] C. Rogers, W.F. Shadwick, Bäcklund Transformations, Academic Press, New York, 1982.
[8] V.A. Matveev, M.A. Salle, Darboux Transformation and Solitons, Springer, Berlin, 1991.
[9] P.J. Olver, Applications of Lie Groups to Differential Equations, second ed., Springer Verlag, New York, 1993.
[10] Mingliang Wang, Solitary wave solutions for variant Boussinesq equations, Physics Letters A 199 (1995) 169–172.
[11] Mingliang Wang, Exact solutions for a compound KdV–Burgers equation, Physics Letters A 213 (1996) 279–287.
[12] Mingliang Wang, Yubin Zhou, Zhibin Li, Application of a homogeneous balance method to exact solutions of nonlinear equations in mathematical
physics, Physics Letters A 216 (1996) 67–75.
[13] Mingliang Wang, Yubin Zhou, The periodic wave solutions for the Klein–Gordon–Schrödinger equations, Physics Letters A 318 (2003) 84–92.
[14] Yubin Zhou, Mingliang Wang, Tiande Miao, The periodic wave solutions and solitary wave solutions for a class of nonlinear partial differential
equations, Physics Letters A 323 (2004) 77–88.
[15] Yubin Zhou, Mingliang Wang, Yueming Wang, Periodic wave solutions to a coupled KdV equations with variable coefficients, Physics Letters A 308
(2003) 31–36.
[16] Abdul-Majid Wazwaz, The extended tan h method for new solitons solutions for many forms of the fifth-order KdV equations, Applied Mathematics
and Computation 184 (2007) 1002–1014.
[17] Abdul-Majid Wazwaz, The extended tan h method for abundant solitary wave solutions of nonlinear wave equations, Applied Mathematics and
Computation 187 (2007) 1131–1142.
[18] Abdul-Majid Wazwaz, The tan hCcot h method for solitons and kink solutions for nonlinear parabolic equations, Applied Mathematics and
Computation 188 (2007) 1467–1475.
[19] Mingliang Wang, Xingzheng Li, Jinliang Zhang, Various exact solutions of nonlinear Schrödinger equation with two nonlinear terms, Chaos, Solitons &
Fractals 31 (2007) 594–601.
[20] Xiangzheng Li, Mingliang Wang, A sub-ODE method for finding exact solutions of a generalized KdV–mKdV equation with high-order nonlinear terms,
Physics Letters A 361 (2007) 115–118.
[21] Mingliang Wang, Xiangzheng Li, Jinliang Zhang, Sub-ODE method and solitary wave solutions for higher order nonlinear Schrödinger equation, Physics
Letters A 363 (2007) 96–101.
[22] Shimin Guo, Yubin Zhou, Auxiliary equation method for the mKdV equation with variable coefficients, Applied Mathematics and Computation (2009),
doi:10.1016/j.amc.2009.06.017.
[23] Abdul-Majid Wazwaz, The sine–cosine method for obtaining solutions with compact and noncompact structures, Applied Mathematics and
Computation 159 (2004) 559–576.
[24] Abdul-Majid Wazwaz, The sine–cosine and the tan h methods: reliable tools for analytic treatment of nonlinear dispersive equations, Applied
Mathematics and Computation 173 (2006) 150–164.
 0
[25] Mingliang Wang, Xiangzheng Li, Jinliang Zhang, The GG -expansion method and travelling wave solutions of nonlinear evolution equations in
mathematical physics, Physics Letters A 372 (2008) 417–423.
 0
[26] Sheng Zhang, Jing-Lin Tong, Wei Wang, A generalized GG -expansion method for the mKdV equation with variable coefficients, Physics Letters A 372
(2008) 2254–2257.  0
[27] Zhou Yu-Bin, Li Chao, Application of modified GG -expansion method to traveling wave solutions for Whitham–Broer–Kaup-Like equations,
Communications in Theoretical Physics51(2009) 664–670.
0
[28] Shimin Guo, Yubin Zhou, The extended GG -expansion method and its applications to the Whitham–Broer–Kaup-Like equations and coupled Hirota–
Satsuma KdV equations, Applied Mathematics and Computation 215 (2010) 3214–3221.
[29] Wen Xiu Ma, Complexiton solutions to the Korteweg–de Vries equation, Physics Letters A 301 (2002) 35–44.
[30] Deng-Shan Wang, Hongbo Li, Symbolic computation and non-travelling wave solutions of (2 + 1)-dimensional nonlinear evolution equations, Chaos,
Solitons & Fractals 38 (2008) 383–390.
[31] Sheng Zhang, A further improved extended Fan sub-equation method for (2 + 1)-dimensional breaking soliton equations, Applied Mathematics and
Computation 199 (2008) 259–267.
[32] Chaoqing Dai, Yueyue Wang, Periodic structures based on variable separation solution of the (2 + 1)-dimensional Boiti–Leon–Pempinelli equation,
Chaos, Solitons & Fractals 39 (2009) 350–355.
[33] Chaoqing Dai, Semifoldons with fusion and fission properties of (2 + 1)-dimensional nonlinear system, Chaos, Solitons & Fractals 38 (2008) 474–480.
[34] Abdul-Majid Wazwaz, Multiple kink solutions and multiple singular kink solutions for the (2 + 1)-dimensional Burgers equations, Applied
Mathematics and Computation 204 (2008) 817–823.
[35] Deng-Shan Wang, Hong-Bo Li, Jike Wang, The novel solutions of auxiliary equation and their application to the (2 + 1)-dimensional Burgers equations,
Chaos, Solitons & Fractals 38 (2008) 374–382.