Chaos, Solitons and Fractals 30 (2006) 700708 www.elsevier.

com/locate/chaos

Exp-function method for nonlinear wave equations

Ji-Huan He *, Xu-Hong Wu
College of Science, Donghua University, 1882 Yan-an Xilu Road, Shanghai 20051, PR China Accepted 7 March 2006

Communicated by Prof. G. Iovane

Abstract In this paper, a new method, called Exp-function method, is proposed to seek solitary solutions, periodic solutions and compacton-like solutions of nonlinear dierential equations. The modied KdV equation and DoddBullough Mikhailov equation are chosen to illustrate the eectiveness and convenience of the suggested method. 2006 Elsevier Ltd. All rights reserved.

1. Introduction Recently many new approaches to nonlinear wave equations have been proposed, for example, tanh-function method [16], F-expansion method [79], Jacobian elliptic function method [1012], variational iteration method [13,14], Adomian method [1518], variational approach [1921], and homotopy perturbation method [2224]. All methods mentioned above have limitation in their applications. In this paper we suggest a novel method called Exp-function method (or Exp-method for short) to search for solitary solutions, compact-like solutions and periodic solutions of various nonlinear wave equations.

2. Basic idea of Exp-function method In order to illustrate the basic idea of the suggested method, we consider rst the following nonlinear dispersive equation of the form [13,2527]: ut u2 ux uxxx 0. 1

This equation is called modied KdV equation, which arises in the process of understanding the role of nonlinear dispersion and in the formation of structures like liquid drops, and it exhibits compactons: solitons with compact support. Introducing a complex variation g dened as g kx xt. 2

Corresponding author. E-mail addresses: jhhe@dhu.edu.cn (J.-H. He), ijnsns@yahoo.com.cn (X.-H. Wu).

0960-0779/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.chaos.2006.03.020

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

701

We have xu0 ku2 u0 k 3 u000 0; 3

where prime denotes the dierential with respect to g. The Exp-function method is very simple and straightforward, it is based on the assumption that traveling wave solutions can be expressed in the following form: Pd an expng ; 4 ug Pqnc bm expmg mp where c, d, p, and q are positive integers which are unknown to be further determined, an and bm are unknown constants. We suppose that the solution of Eq. (3) can be expressed as ug ac expcg ad expdg . ap exppg aq expqg 5

To determine values of c and p, we balance the linear term of highest order in Eq. (3) with the highest order nonlinear term. By simple calculation, we have u000 and u2 u0 c3 expp 3cg c3 exp5p 3cg ; c4 exp4pg c4 exp8pg 7 c1 exp7p cg c2 exp8pg 6

where ci are determined coecients only for simplicity. Balancing highest order of Exp-function in Eqs. (6) and (7), we have 7p c 5p 3c; which leads to the result p c. Similarly to determine values of d and q, we balance the linear term of lowest order in Eq. (3) u000 and u2 u0 d 3 expq 3dg d 3 exp5q 3dg ; d 4 exp4qg d 4 exp8qg 11 d 1 exp7q dg d 2 exp8qg 10 9 8

where di are determined coecients only for simplicity. Balancing lowest order of Exp-function in Eqs. (10) and (11), we have 7q d 5q 3d; which leads to the result q d. For simplicity, we set p = c = 1 and q = d = 1, so Eq. (5) reduces to ug a1 expg a0 a1 expg . expg b0 a1 expg 14 13 12

Substituting Eq. (14) into Eq. (3), and by the help of Matlab, we have 1 C 3 exp3g C 2 exp2g C 1 expg C 0 C 1 expg C 2 exp2g C 3 exp4g 0; A 15

702

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

where A expg b1 expg b0 4 ; C 3 xa1 b0 ka3 b0 k 3 a0 wa0 ka2 a0 k 3 a1 b0 ; 1 1 C 2 8k 3 a1 b1 2ka3 b1 4k 3 a1 b2 2wa1 2ka1 a2 2wa1 b1 4k 3 a0 b0 2ka2 a1 1 0 1 0 2ka2 a0 b0 2xa1 b2 2xa0 b0 8k 3 a1 ; 1 0 C 1 xa1 b3 6xa1 b0 b1 xa0 b2 k 3 a0 b2 18k 3 a1 b0 b1 6ka1 a0 a1 ka1 a2 b0 ka3 0 0 0 0 0 23k 3 a0 b1 xa0 b1 5xa1 b0 k 3 a1 b3 5k 3 a1 b0 ka2 a1 b0 5ka2 a0 b1 ; 0 1 1 C 0 4xa1 b2 4ka1 a2 32k 3 a1 b1 4ka1 a2 b1 32k 3 a1 b2 4k 3 a1 b2 b1 4xa1 b1 1 1 0 1 0 4k 3 a1 b2 4ka2 a1 4xa1 b2 4ka2 a1 b1 4xa1 b2 b1 ; 0 0 0 1 0 C 1 18k 3 a1 b0 b1 6xa1 b0 b1 k 3 a1 b3 k 3 a0 b1 b2 xa0 b2 5ka0 a2 5xa1 b0 b2 0 0 1 1 1 xa0 b1 b2 xa1 b3 ka1 a2 b0 23k 3 a0 b2 ka2 a1 b0 5k 3 a1 b0 b2 ka3 b1 1 0 0 1 0 1 0 6ka1 a0 a1 b1 ; C 2 2xa0 b2 b0 2xa1 b2 2ka3 2ka1 a2 b1 2xa1 b3 4k 3 a0 b2 b0 2xa1 b2 b1 1 1 1 1 1 1 0 4k 3 a1 b2 b1 8k 3 a1 b2 2ka2 a1 b1 2ka0 a2 b0 8k 3 a1 b3 ; 1 0 1 0 1 C 3 ka0 a2 b1 xa0 b3 ka3 b0 k 3 a0 b3 xa1 b0 b2 k 3 a1 b0 b2 . 1 1 1 1 1 1 Equating the coecients of exp(ng) to be zero, we have 8 > C 3 0; C 2 0; C 1 0; < C 0; > 0 : C 2 0; C 3 0; C 4 0: Solving the system, Eq. (16), simultaneously, we obtain 8 2 2 2 2 < a0 a1 b0 3ka b0 ; a1 b0 3k 2a1 ; 8a1 1 :b
1

16

b2 3k 2 2a2 0 1 ; 8a2 1

x ka2 k 3 ; 1

17

where a1 and b0 are free parameters. We, therefore, obtain the following solution: ux; t a1 expkx ka2 k 3 t a1 b0 3ka1b0 1 expkx ka2 k 3 t b0 1
8a2 1
2

b2 3k 2 2a2 0 1 8a1

expkx ka2 k 3 t 1

b2 3k 2 2a2 0 1

expkx ka2 k 3 t 1 . 18

a1

3k 2 b0 a1

expkx ka2 k 3 t b0 1

b2 3k 2 2a2 0 1 8a2 1

expkx ka2 k 3 t 1

Generally a1, b0, and k are real numbers, and the obtained solution, Eq. (18), is a generalized solitonary solution. In case k is an imaginary number, the obtained solitonary solution can be converted into periodic solution or compact-like solution. We write k iK. Use the transformation expkx ka2 k 3 t expiKx iKa2 K 3 t cosKx Ka2 K 3 t i sinKx Ka2 K 3 t 1 1 1 1 and expkx ka2 k 3 t expiKx iKa2 K 3 t cosKx Ka2 K 3 t i sinKx Ka2 K 3 t. 1 1 1 1 Eq. (18) becomes ux; t a1 where p 3K 1b0 a 1 p cosKx Ka2 K 3 t b0 i1 p sinKx Ka2 K 3 t 1 1
2

19

20

b2 3k 2 2a2 0 1 . 8a2 1

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

703

If we search for a periodic solution or compact-like solution, the imaginary part in the denominator of Eq. (20) must be zero, that requires that 1p 1 b2 3K 2 2a2 0 1 0. 8a2 1 21

Solving b0 from Eq. (21) we obtain s 8 . b0 3K 2 2a2 1 Substituting Eq. (22) into Eq. (20) results in a compact-like solution, which reads q q 3K 2 3K 282a2 3K 2 3K 222a2 1 1 q a1 q ; ux; t a1 2 cosKx Ka2 K 3 t 3K 282a2 a1 cosKx Ka2 K 3 t 3K 222a2 a1 1 1
1 1

22

23

where a1 and K are free parameters, and it requires that 2a2 > 3K 2 . 1 To compare our result, Eq. (23), with that in open literature, we write down Zhu et al.s solutions [28,29], which read p 4 2k sin2 kx 4k 3 t ux; t . 24a 3 2 sin2 kx 4k 3 t and ux; t p 4 2k cos2 kx 4k 3 t . 3 2 cos2 kx 4k 3 t 24b

We re-write Eqs. (24a) and (24b), respectively, in the forms p p 6 2k ux; t 2 2k 2 cos2kx 8k 3 t and p ux; t 2 2k p 6 2k . 2 cos2kx 8k 3 t

25a

25b

p If we choose a1 2 2k, our solution, Eq. (23), turns out to be Zhu et al.s solutions as expressed in Eqs. (25a) and (25b). So the suggested Exp-function method can obtain easily the generalized solitonary solution and compact-like solution for nonlinear wave equations. To illustrate its eectives and convenience, we consider in the next section the Dodd BulloughMikhailov equation. 3. An example Now we consider the DoddBulloughMikhailov equation [4] uxt eu e2u 0. 26

This equation plays a signicant role in many scientic applications such as solid state physics, nonlinear optics and quantum eld theory. By the transformation u = ln v, Eq. (26) becomes vvxt vx vt v3 1 0. Introducing a complex variation g dened as g = kx + xt, we have f v  kxvv00 kxv0 2 v3 1 0; where prime denotes the dierential with respect to g. We suppose that the solution of Eq. (28), can be expressed as ux; t ac expckx xt ad expdkx xt . ap exppkx xt aq expqkx xt 29 28 27

704

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

By the same manipulation as illustrated in the previous section, we can determine values of c and p by balancing vv00 and v3 in Eq. (28) vv00 and v3 c3 exp3cg c1 exp2p 3cg . c4 exp3pg c4 exp5pg 31 c1 exp3p 2cg c2 exp5pg 30

Balancing the highest order of Exp-function in Eqs. (30) and (31), we have 3p 2c 2p 3c; which leads to the result p c. By a similar derivation as illustrated in the previous section, we obtain d q. 3.1. Case 1: p = c = 1, d = q = 1 We can freely choose the values of c and d, but we will illustrate that the nal solution does not strongly depends upon the choice of values of c and d. For simplicity we choose p = c = 1 and d = q = 1, the trial-function, Eq. (29), becomes vg a1 expg a0 a1 expg . expg b0 a1 expg 35 34 33 32

Substituting (35) into (28), by help of Matlab, we have 1 fC 4 exp4g C 3 exp3g C 2 exp2g C 1 expg C 0 C 1 expg C 2 exp2g C 3 exp3g A C 4 exp4gg 0; 36 where A expg b1 expg b0 4 ; C 4 a3 1; C 3 3a2 a0 kxa0 a1 kxa2 b0 a3 b0 4b0 ; 1 1 1 C 2 3a1 a2 4b1 4kxa1 a1 6b2 4kxa2 b1 a3 b1 3a2 a0 b0 3a2 a1 ; 0 1 1 1 1 0 C 1 a3 6kxa1 b1 a0 6kxa1 b0 a1 12b0 b1 kxa2 b0 kxa2 b0 b1 6a1 a0 a1 0 0 1 kxa0 a1 b2 3a1 a2 b0 4b3 3a2 a0 b1 3a2 a1 b0 kxa0 a1 ; 0 1 1 0 0 C 0 3a2 a1 b1 6b2 12b2 b1 4kxa2 b1 3a1 a2 3a2 a1 6a1 a0 a1 b0 b4 1 0 1 0 1 0 0 4kxa1 a1 b2 a3 b0 3a1 a2 b1 ; 0 0 0 C 1 a3 b1 3a0 a2 3a2 a1 b0 kxa0 a1 b2 6kxa1 b0 a1 b1 kxa0 b2 a1 4b3 b1 0 1 0 0 1 0 kxa2 b1 b0 6a1 a0 a1 b1 6kxa0 a1 b1 kxa2 b0 3a1 a2 b0 12b0 b2 ; 1 0 1 1 C 2 6b2 b2 4kxa2 b1 a3 3a2 a1 b1 3a1 a2 b1 4b3 4kxa1 a1 b2 3a0 a2 b0 ; 0 1 1 1 1 1 0 1 1 C 3 3a0 a2 b1 kxa2 b0 b1 a3 b0 4b0 b3 kxa1 a0 b3 ; 1 1 1 1 1 C 4 b4 a3 b1 . 1 1 Equating the coecients of exp(ng) in Eq. (36) to be zero, we have 8 > C 4 0; C 3 0; C 2 0; C 1 0; < C 0; > 0 : C 1 0; C 2 0; C 3 0; C 4 0:

37

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

705

Solving the system, Eq(37), simultaneously yields b2 b2 0 ; b1 0 ; a1 1; a0 2b0 ; a1 4 4

3 x . k

38

We, therefore, obtain b2 exp kx 3 t 2b0 40 exp kx 3 t k k vx; t ; b2 exp kx 3 t b0 40 exp kx 3 t k k

39

where k and b0 are non-zero free parameters The solution of the DoddBulloughMikhailov equation can be expressed as follows: " # 3b0 ux; t ln v ln 1 40 k k 2 . 3 3 exp 2 x 2k t b20 exp 2 x 2k t p p p p 3 Hereby it requires that b0 > 0 and ln 321 b0 < k x 2k t < ln 321 b0 . 2 To compare our result with the known ones in literature, we write down Wazwazs solution [4], which reads " r #! 1 3 2 1 1 3tanh vx; t x ct 41 2 2 c We re-write it in the form 0 2  q   q 32 1 p p 1 3 1 1 3 1 1B 6exp 2 c x 2 3c t exp 2 c x 2 3c t 7 C vx; t @1 3 4  q   q 5 A p p 2 exp 1 3 x 1 3c t exp 1 3 x 1 3c t 2 c 2 2 c 2 q  q p  p  exp 3x 3ct 4 exp 3x 3ct c c q  q p  p  : exp 3x 3ct 2 exp 3x 3ct c c

42

It is obvious that in case b0 = 2, our solution reduces to Wazwazs. As illustrated above that the obtained solitonary solution can be converted into periodic solution or compact-like solution if k is chosen as an imaginary number. If k = iK, then Eq. (39) becomes     b2 b2 3 3 1 40 cos Kx K t i 1 40 sin Kx K t 2b0   vx; t  . 43  b2 b2 3 3 1 40 cos Kx K t i 1 40 sin Kx K t b0 Elimination of the imaginary part requires that b0 2. Eq. (43) reduces to 3 cos Kx K t 2 3 vx; t 1 . 3 3 cos Kx K t 1 cos Kx K t 1 44

45

We, therefore, obtain a periodic solution, which reads ! 3 ux; t ln v ln 1 ; 3 cos Kx K t 1

3 it requires that Kx K t 6 2n 1p.

46

3.2. Case 2: p = c = 2, d = q = 2 As mentioned above the values of c and d can be freely chosen, now we set p = c = 2 and d = q = 2, then the trialfunction, Eq. (29), becomes vg a2 exp2g a1 expg a0 a1 expg a2 exp2g . exp2g b1 expg b0 b1 expg b2 expg 47

706

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

There are some free parameters in Eq. (47), we set b1 = b1 = 0 for simplicity, the trial-function (47) is simplied as follows: vg a2 exp2g a1 expg a0 a1 expg a2 exp2g . exp2g b0 b2 expg 48

By the same manipulation as illustrated above, we obtain 8 < a2 1; a0 5 a2 ; 18 1 : b a4 ; x 3 ; 1 2 1296 k or 8 < a2 1; : b b2 ; 0

2 4
1 a1 36 ;

a3

1 a2 1296 ;

a4

49a

a1 0;
3 x 4k :

a0 2b0 ;

a1 0;

a2 40 ;

b2

49b

Eq. (49b) leads to the following solution: 5 a3 a4 1 1 exp 2kx 6 t a1 exp kx 3 t 18 a2 36 exp kx 3 t 1296 2kx 6 t 1 k k k k . vx; t a2 a4 1 1 exp 2kx 6 t 18 1296 2kx 6 t k k Setting a1 = 6a, Eq. (50) is simplied as follows: exp 2kx 6 t 6a exp kx 3 t 10a2 6a3 exp kx 3 t a4 2kx 6 t k k k k vx; t exp 2kx 6 t 2a2 a4 2kx 6 t k k exp kx 3 t a2 exp kx 3 t 2a exp kx 3 t a2 exp kx 3 t 4a k k k k 2 exp kx 3 t a2 exp kx 3 t 4a2 k k exp kx 3 t a2 exp kx 3 t 4a k k . exp kx 3 t a2 exp kx 3 t 2a k k Substituting Eq. (49b) into Eq. (48) results in another solution, which reads vx; t b2 exp kx 3 t 2b0 40 exp kx 3 t k k . b2 exp kx 3 t b0 40 exp kx 3 t k k 51b 50

51a

It is interesting that we nd the same solutions as those in the Case 1. 3.3. Case 3: p = c = 2, d = q = 1 Now we consider the case p = c = 2 and d = q = 1. Under such case, the trial-function can be expressed as follows: vg a2 exp2g a1 expg a0 a1 expg . exp2g b1 expg b0 b1 expg 52

By simple calculation by Matlab, we have 8 2 > a2 1; a0 a1 a1 b1 5 b2 ; > > 4 6 12 1 <

a1 b1 1 > b0 12 6 > > : x 3: k a2 b2 1 4

1 a1 108

a3

b2 a1 1 36

1 54 ;

b3

1 b1 108

a3

b2 a1 1 36

1 54 ;

b3

53

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

707

So the solution can be expressed as  3  a2 a1 b3 b2 a 5 1 1 exp 2kx 6 t a1 exp kx 3 t 41 a16b1 12 b2 108 361 54 exp kx 3 t 1 k k k  3  vx; t a2 a1 b3 b2 a 1 1 1 exp 2kx 6 t b1 exp kx 3 t 12 a16b1 1 b2 108 361 54 exp kx 3 t 1 k k 4 k a2 a1 b1 exp kx 3 t 31 a13b1 2 b2 k  3 3 21  1 a2 a1 b3 b1 a 1 1 exp 2kx 6 t b1 exp kx 3 t 12 a16b1 1 b2 108 361 54 exp kx 3 t k k 4 1 k 36a1 b1 3 exp kx 3 t a1 2b1 k 1 2 exp kx 3 t 3 exp kx 3 t a1 2b1 6 exp kx 3 t a1 b1 k k k 36a1 b1 exp kx 3 t k 1 2 . 6 exp kx 3 t a1 b1 k If we set 6c = a1 + b1, Eq. (54) reduces to 6c exp kx 3 t 6c k ; vx; t 1 2 1 exp kx 3 t 2c c2 exp kx 3 t exp kx 3 t c k k k which is same as those obtained in Cases 1 and 2.

54

55

4. Conclusion We give a very simple and straightforward method called Exp-function method for nonlinear wave equations. The suggest method has some pronounced merits: (1) The method leads to both the generalized solitonary solutions and periodic solutions; (2) The solution procedure, by help of Matlab, is of utter simplicity, and can be easily extended to all kinds of nonlinear equations. The Exp-function method might become a promising and powerful new method for nonlinear equations.

Acknowledgement This work is supported by Program for New Century Excellent Talents in University and the William M.W. Wong Engineering Research Fund.

References
[1] Abdusalam HA. On an improved complex tanh-function method. Int J Nonlinear Sci Numer Simul 2005;6(2):99106. [2] Zayed EME, Zedan HA, Gepreel KA. Group analysis and modied extended tanh-function to nd the invariant solutions and soliton solutions for nonlinear Euler equations. Int J Nonlinear Sci Numer Simul 2004;5(3):22134. [3] Bai CL, Zhao H. Generalized extended tanh-function method and its application. Chaos, Solitons & Fractals 2006;27(4):102635. [4] Wazwaz AM. The tanh method: solitons and periodic solutions for the DoddBulloughMikhailov and the TzitzeicaDodd Bullough equations. Chaos, Solitons & Fractals 2005;25(1):5563. [5] Li DS, Gao F, Zhang HQ. Solving the (2 + 1)-dimensional higher order BroerKaup system via a transformation and tanhfunction method. Chaos, Solitons & Fractals 2004;20(5):10215. [6] Khuri SA. A complex tanh-function method applied to nonlinear equations of Schroedinger type. Chaos, Solitons & Fractals 2004;20(5):103740. [7] Yomba E. The modied extended Fan sub-equation method and its application to the (2 + 1)-dimensional BroerKaup Kupershmidt equation. Chaos, Solitons & Fractals 2006;27(1):18796. [8] Ren YJ, Zhang HQ. A generalized F-expansion method to nd abundant families of Jacobi elliptic function solutions of the (2 + 1)-dimensional NizhnikNovikovVeselov equation. Chaos, Solitons & Fractals 2006;27(4):95979. [9] Wang DS, Zhang HQ. Further improved F-expansion method and new exact solutions of KonopelchenkoDubrovsky equation. Chaos, Solitons & Fractals 2005;25(3):60110.

708

J.-H. He, X.-H. Wu / Chaos, Solitons and Fractals 30 (2006) 700708

[10] Dai CQ, Zhang JF. Jacobian elliptic function method for nonlinear dierentialdierence equations. Chaos, Solitons & Fractals 2006;27(4):10427. [11] Yu YX, Wang Q, Zhang HQ. The extended Jacobi elliptic function method to solve a generalized HirotaSatsuma coupled KdV equations. Chaos, Solitons & Fractals 2005;26(5):141521. [12] Zhao XQ, Zhi HY, Zhang HQ. Improved Jacobi-function method with symbolic computation to construct new double-periodic solutions for the generalized Ito system. Chaos, Solitons & Fractals 2006;28(1):11226. [13] He JH, Wu XH. Construction of solitary solution and compacton-like solution by variational iteration method. Chaos, Solitons & Fractals 2006;29(1):10813. [14] Momani S, Abuasad S. Application of Hes variational iteration method to Helmholtz equation. Chaos, Solitons & Fractals 2006;27(5):111923. [15] Abassy TA, El-Tawil MA, Saleh HK. The solution of KdV and mKdV equations using Adomian Pade approximation. Int J Nonlinear Sci Numer Simul 2004;5(4):32739. [16] El-Sayed SM, Kaya D, Zarea S. The decomposition method applied to solve high-order linear VolterraFredholm integrodierential equations. Int J Nonlinear Sci Numer Simul 2004;5(2):10512. [17] El-Danaf TS, Ramadan MA, Alaal FEIA. The use of adomian decomposition method for solving the regularized long-wave equation. Chaos, Solitons & Fractals 2005;26(3):74757. [18] Zhu YG, Chang QS, Wu SC. Construction of exact solitary solutions for Boussinesq-like B(m, n) equations with fully nonlinear dispersion by the decomposition method. Chaos, Solitons & Fractals 2005;26(3):897903. [19] Liu HM. Generalized variational principles for ion acoustic plasma waves by Hes semi-inverse method. Chaos, Solitons & Fractals 2005;23(2):5736. [20] Liu HM. Variational approach to nonlinear electrochemical system. Int J Nonlinear Sci Numer Simul 2004;5(1):956. [21] Zhang J, Yu JY, Pan N. Variational principles for nonlinear ber optics. Chaos, Solitons & Fractals 2005;24(1):30911. [22] He JH. Application of homotopy perturbation method to nonlinear wave equations. Chaos, Solitons & Fractals 2005;26(3): 695700. [23] He JH. Homotopy perturbation method for bifurcation of nonlinear problems. Int J Nonlinear Sci Numer Simul 2005;6(2):2078. [24] El-Shahed M. Application of Hes Homotopy perturbation method to Volterras integro-dierential equation. Int J Nonlinear Sci Numer Simul 2005;6(2):1638. [25] Rosenau P, Hyman JM. Compactons: solitons with nite wavelengths. Phys Rev Lett 1993;75(5):5647. [26] Rosenau P. Compact and noncompact dispersive patterns. Phys Lett A 2000;275(3):193203. [27] Wazwaz AM. Two reliable methods for solving variants of the KdV equation with compact and noncompact structures. Chaos, Solitons & Fractals 2006;28(2):45462. [28] Zhu Y, Gao X. Exact special solitary solutions with compact support for the nonlinear dispersive K(m, n) equations. Chaos, Solitons & Fractals 2006;27(2):48793. [29] Zhu Y, Chang Q, Wu S. Exact solitary-wave solutions with compact support for the modied KdV equation. Chaos, Solitons & Fractals 2005;24(1):3659.