The Riccati Equation Method With Variable Expansion Coefficients. IV. Solving The Fitzhugh-Nagumo Equation.

The Riccati equation method
with variable expansion coefficients.

IV. Solving the Fitzhugh!agumo equation.

Solomon M. Antoniou
SKEMSYS
Scientific Knowledge Engineering
and Management Systems

Corinthos 20100, Greece
solomon_antoniou@yahoo.com

"bstract
The Riccati equation method with variable expansion coefficients, introduced in
previous papers, is used to find traveling wave solutions to the Fitzhugh-Nagumo
(FN) equation ) u a ( ) u 1 ( u u u
xx t
= . The dependent coefficients A and
B of the Riccati equation
2
BY A Y + = are either proportional each other or their
product is an exponential function. They are determined as solutions of ODEs they
satisfy and their solutions are expressed either in terms of Bessel's functions or in
terms of functions already found in Paper I. The same situation occurs for the
expansion coefficients. The function Y is calculated to be either a constant or it
can be expressed in terms of Bessel's functions.

Keywords: Riccati equation method, nonlinear evolution equations, traveling
wave solutions, Fitzhugh-Nagumo equation, Bessel functions, exact solutions.

2

1. Introduction.
Nonlinear partial differential equations arise in a number of areas of Mathematics
and Physics in an attempt to model physical processes, like Chemical Kinetics
(Gray and Scott [42]), Fluid Mechanics (Whitham [97]), or biological processes
like Population Dynamics (Murray [67]). In the recent past there are a number of
new methods which have been invented in solving these equations. Among the
new methods are the inverse scattering method (AKNS [12], Ablowitz and Segur
[14], Ablowitz and Clarkson [13], Novikov, Manakov, Pitaevskii and Zakharov
[72]), Hirotas bilinear method (Hirota [48] and [49]), the algebro-geometric
approach (Belokolos et al [21]), the tanh-coth method (Malfliet [63] and [64],
Malfliet and Hereman [65] and [66], El-Wakil and Abdou [29], Fan [34], Griffiths
and Sciesser [43], Fan and Hon [35], Parkes and Duffy [77], Parkes, Zhou, Duffy
and Huang [79], Wazwaz [92] ), the sn-cn method (Baldwin et al [19]), the F-
expansion method (Abdou [5] and [8], Wang and Li [90]), the Jacobi elliptic
function method (Abbott, Parkes and Duffy [2], Abdou and Elhanbaly [11], Chen
and Zhang [24], Chen and Wang [25], Fan and Zhang [36], Inc and Ergt [51],
Liu, Fu, Liu and Zhao [60], [61], Lu and Shi [62], Parkes, Duffy and Abbott [78]),
the Riccati equation method (Zhang and Zhang [103], Abdou [4], Antoniou [16]
and [17]), the Weierstrass elliptic function method (Kudryashov [55], [57]), the
exp-function method (He and Wu [46], Abdou [9], Aslan [18], Bekir and Boz
[20], Ebaid [27], El-Wakil, Abdou and Hendi [30], He and Abdou [45], Naher,
Abdullah and Akbar [69] and [70]), the Bcklund transformation method (Rogers
and Shadwick [82]), the G / G expansion method (Borhanibar and Moghanlu
[23], Feng, Li and Wan [37], Jabbari, Kheiri and Bekir [52], Naher, Abdullah and
Akbar [71], Ozis and Aslan [76], Wang, Li and Zhang [91], Zayed [100], Zayed
and Gepreel [102], Antoniou [16] and [17]), the homogeneous balance method
(Fan [33], Wang, Zhou and Li [89], El-Wakil, Abulwafa, Elhanbaly and Abdou
[28], Wang, Zhou and Li [89]), the direct algebraic method (Soliman and Abdou
3
[85]), the basic equation method (Kudryashov [56]) and its variants, like the
simplest equation method (Abdou [7], Jawad, Petkovich and Biswas [53], Vitanov
[88], Yefimova [99], Zayed [101]), the first integral method (Feng [38], Raslan
[80]), the integral bifurcation method (Rui, Xie, Long and He [83]), the reduced
differential transform method (Keskin and Oturanc [54]), the Cole-Hopf
transformation method (Salas and Gomez [84]), the Adomian decomposition
method (Adomian [15], Abdou [3], Wazwaz [92] and [94]), the Painleve
truncated method (Weiss, Tabor and Carnevale [95] and [96]), the homotopy
perturbation method (Taghizadeh, Akbari and Ghelichzadeh [87], Yahya et al
[98], Liao [59], El-Wakil and Abdou [31]), the Lie symmetry method (Lie point
symmetries, potential symmetries, nonclassical symmetries, the direct method)
(Bluman and Kumei [22], Hydon [50], Olver [74], Ovsiannikov [75], Stephani
[86]), the variational iteration method (He [44], Abdou [6], Abdou and Soliman
[10], Wazwaz [94]). A more detailed, although not complete set of references of
the above methods, appears in Antoniou [16].
The implementation of most of these methods was made possible only using
Symbolic Languages like Mathematica, Macsyma, Maple, etc.
In this paper we implement the Riccati equation method with variable expansion
coefficients introduced previously (Antoniou [16], hereafter called Paper I), and
we find traveling wave solutions of the Fitzhugh-Nagumo equation.
The paper is organized as follows: In Section 2 we introduce the basic ingredients
of the method used. In Section 3 we consider Fitzhugh-Nagumo equation and
Riccatis equation method of solution where the expansion coefficients depend on
the variable . We find that the dependent coefficients A and B of the Riccati
equation
2
BY A Y + = are either proportional each other or their product is an
exponential function and satisfy their own nonlinear ODEs. Their solutions are
expressed either in terms of Bessel's functions or in terms of functions already
found in Paper I. The expansion coefficients are expressed either in terms of
4
Bessel's functions or in terms of the function ) ( A A . The solution Y of
Riccati's equation is expressed either in terms of the proportionality factor or in
terms of Bessel's functions. In Section 4 we consider the
|
\
|

G
G
method with
variable expansion coefficients.

2. The Method.
We consider an evolution equation of the general form
) , u , u , u ( G u
xx x t
L = or ) , u , u , u ( G u
xx x tt
L = (2.1)
where u is a sufficiently smooth function. We introduce a new variable given
by
) t x ( k = (2.2)
where k and are constants. Changing variables, since
) ( u ) k ( u
t
= , ) ( u k u
x
= , ) ( u k u
2
xx
= , (2.3)
equation (2.1) becomes an ordinary nonlinear differential equation

|
|
\
|
L ,
d
u d
k ,
d
du
k , u G
d
du
) k (
2
2
2
(2.4)
or
|
|
\
|
L ,
d
u d
k ,
d
du
k , u G
d
u d
k
2
2
2
2
2
2
(2.5)
Equations (2.4) or (2.5) can be solved considering expansions of the form
=
=
n
0 k
k
k
Y a ) ( u (2.6)
or

= =
+ =
n
1 k
k
k
n
0 k
k
k
Y
b
Y a ) ( u (2.7)
where all the expansion coefficients depend on the variable ,
5
) ( a a
k k
, ) ( b b
k k
for every n , , 2 , 1 , 0 k L =
contrary to the previously considered cases where the expansion coefficients were
considered as constants. The function ) ( Y Y satisfies Riccatis equation

2
Y B A ) ( Y + = (2.8)
where again the coefficients A and B depend on the variable .
In solving equations (2.4) or (2.5), we consider the expansions (2.6) or (2.7) and
then we balance the nonlinear term with the highest derivative of the function
) ( u which determines n (the number of the expansion terms). Equating similar
powers of the function ) ( Y we can determine the various coefficients and thus
find the solution of the equation considered.
In our two previous papers (Antoniou [16] and [17]) we have applied successfully
this method into two notable equations, the Burgers equation and the KdV
equation. We have also considered the accompanied ) G / G ( method with
variable expansion coefficients which can only be applied provided the Riccati
equation method has applied before to the same equation. A remarkable feature of
the method is that we obtain quite new solutions and apart of that, we get a
proliferation of solutions, once we continue to apply the method to all the
intermediate equations.

3. The FN equation and its solutions.
The FitzHugh-Nagumo equation is the mathematical implementation of the
Hodgkin-Buxley model of nerve membranes. It was introduced by R. FitzHugh
(FitzHugh [39], [40] and [41]) and by Nagumo (Nagumo, Arimoto and Yoshizawa
[68]) who also constructed the equivalent circuit. The FN equation can be solved
through a number of methods, like the homotopy analysis method (Abbasbandy
[1]), the Cole-Hopf quotient method (Dengyuan and Yuan [26]), the nonclassical
method (Nucci and Clarkson [73], Estevez [32]), and the first integral method (Li
6
and Guo [58]). The qualitative theory of the equation is best described by J.
Rauch and J. Smoller (Rauch and Smoller [81]).
In this paper we consider the FN equation in the form
) u a ( ) u 1 ( u u u
xx t
= (3.1)
and try to find traveling wave solutions of this equation. We introduce a new
variable given by
t x = (3.2)
where is a constant. Changing variables, since
) ( u ) ( u
t
= , ) ( u u
x
= and ) ( u u
xx
= (3.3)
equation (3.1) becomes an ordinary nonlinear differential equation
0 ) ( u ) ( u ) a 1 ( ) ( u a ) ( u ) ( u
3 2
= + + + (3.4)
We consider the extended Riccati equation method in solving equation (3.4), since
the Riccati equation method leads to trivial results. In this case we consider the
expansion

= =
+ =
n
1 k
k
k
n
0 k
k
k
Y
b
Y a ) ( u
and balance the second order derivative term with the third order nonlinear term of
(3.4). We then find 1 n = and thus

Y
b
Y a a ) ( u
1
1 0
+ + = (3.5)
where again all the coefficients
0
a ,
1
a and
1
b depend on , and Y satisfies
Riccatis equation
2
BY A Y + = . From equation (3.5) we obtain (taking into
account
2
BY A Y + = )

2
1 1 2
1 1 1 1 0
Y
A b
Y
b
Y B a Y a ) A a B b a ( ) ( u
+ + + + = (3.6)
+ + + = ) B b 2 B b A a A a 2 a ( ) ( u
1 1 1 1 0

+ + + + + +
3 2
1
2
1 1 1 1
Y ) B a 2 ( Y ) B a 2 B a ( Y ) B A a 2 a (
7
3
2
1
2
1 1 1 1
Y
A b 2
Y
A b A b 2
Y
B A b 2 b
+
+
+
+ (3.7)
Therefore equation (3.4), under the substitution (3.5), (3.6) and (3.7), becomes
+ + + ) B b 2 B b A a A a 2 a (
1 1 1 1 0

+ + + + + +
3 2
1
2
1 1 1 1
Y ) B a 2 ( Y ) B a 2 B a ( Y ) B A a 2 a (
3
2
1
2
1 1 1 1
Y
A b 2
Y
A b A b 2
Y
B A b 2 b
+
+
+
+
|
\
|
+ +
)
`
+ + + + +
Y
b
Y a a a
Y
A b
Y
b
Y B a Y a ) A a B b a (
1
1 0
2
1 1 2
1 1 1 1 0

0
Y
b
Y a a
Y
b
Y a a ) a 1 (
3
1
1 0
2
1
1 0
=
|
\
|
+ +
|
\
|
+ + + + (3.8)
Upon expanding the previous equation and equating to zero the coefficients of Y,
we obtain a system of ordinary differential equations from which we can
determine the various expansion coefficients. We obtain
coefficient of
3
Y :
0 B a 2 a
2
1
3
1
= + (3.9)
coefficient of
2
Y :
0 B a B a a a 3 a ) a 1 ( B a 2
1 1
2
1 0
2
1 1
= + + + + (3.10)
coefficient of Y:
0 ) a a a b ( 3 a a a B A a 2 a a a ) a 1 ( 2
2
0 1
2
1 1 1 1 1 1 1 0
= + + + + + (3.11)
coefficient of
0
Y :
B b 2 A a B b A a 2 ) a b a 2 ( ) a 1 ( a a a
1 1 1 1
2
0 1 1 0 0
+ + + + +
0 ) b a 6 a ( a ) B b A a a (
1 1
2
0 0 1 1 0
= + + + (3.12)
coefficient of
1
Y
:
0 ) b a b a ( 3 b b a ) a 1 ( 2 B A b 2 b a b
2
1 1 1
2
0 1 1 0 1 1 1
= + + + + + (3.13)
8
coefficient of
2
Y
:
0 A b 2 A b b ) a 1 ( b A b a 3
1 1
2
1 1
2
1 0
= + + (3.14)
coefficient of
3
Y

:
0 b A b 2
3
1
2
1
= (3.15)
The system of equations (3.9)-(3.15) are to be solved supplemented by the Riccati
equation
2
BY A Y + = .
From equations (3.9) and (3.15), ignoring the trivial solutions, we obtain
B 2 a
1
= and A 2 b
1
= (3.16)
respectively. We thus consider the next four cases.
3.I. Case I. For B 2 a
1
= and A 2 b
1
= , we obtain
from (3.14)

A
A
2 3 ) a 1 ( 2 2 a 6
0
+ + = (3.17)
from (3.10)

B
B
2 3 ) a 1 ( 2 2 a 6
0
+ + + = (3.18)
from (3.11)
0 a a ) a 1 ( 2 a 3 AB 4
B
B
B
B
0
2
0
= + +
+

(3.19)
from (3.13)
0 a a ) a 1 ( 2 a 3 AB 4
A
A
A
A
0
2
0
= + +
+

(3.20)
from (3.12)

3
0
2
0 0 0 0 0
a a ) a 1 ( B A ) a 1 ( 4 B A a 12 a a a a + + + + +
0 ) B A B A ( 2 = + (3.21)
Equating the two different expressions of
0
a , equations (3.17) and (3.18), we
9
obtain the equation =
|
\
|

+
2
B
B
A
A
3 , which by integration gives

|
\
|

=
3
2
exp K B A (3.22)
where K is a constant.
Using (3.18) and (3.22), we obtain from (3.19) the equation
0 ) 2 a 2 a 2 (
6
1
3
2
exp K 4
B
B
2
3
B
B
2 2
2
= + +
|
\
|

\
|

(3.23)
The previous equation, under the substitution

B
B
F

= (3.24)
transforms into the equation
) a 2 a 2 2 (
6
1
3
2
exp K 4 F
2
1
F
2 2 2
+
|
\
|

+ = (3.25)
which is a Riccati differential equation. Under the standard transformation

w
w
2 F

= (3.26)
equation (3.25) takes on the form of a second order differential equation
0 w ) a 2 a 2 2 (
12
1
3
2
exp K 2 w
2 2
=
)
`
+
|
\
|

+ (3.27)
The substitution

= e z transforms (3.27) into
0 w ) a 2 a 2 2 (
12
1
z K 2
1
dz
dw
z
dz
w d
z
2 2 3
2
2 2
2
2
=
+ +

(3.28)
The choice
3
= converts
3
2
z into
2
z and thus equation (3.28) takes on
the form
0 w ) a 2 a 2 2 (
4
3
z
K 18
dz
dw
z
dz
w d
z
2 2
2
2
2 2
2
2
=
+ + (3.29)
10
The substitution

2
2
K 18
= and ) a 2 a 2 2 (
4
3
2 2
2
2
+
= (3.30)
converts (3.29) into
0 w ) z (
dz
dw
z
dz
w d
z
2 2 2
2
2
2
= + + (3.31)
which is Bessel's differential equation with general solution
) z ( Y C ) z ( J C w
2 1
+ =

(3.32)
It is apparent from (3.30) that the constant K should be positive: 0 K > .
Going back to the original variables, we have

|
|
\
|
+
|
|
\
|
3
2
3
1
e
K 2 3
Y C e
K 2 3
J C ) ( w (3.33)
where

2 2
3 ) 1 a a ( 6
2
1
+
= (3.34)
with 0 ) 1 a a ( 2
2 2
> + .
Combining (3.24) and (3.26), we derive the equation 0
w
w
2
B
B
=
, from which
by integration we obtain
) ( w
C
) ( B
2
= (3.35)
with w given by (3.33). Using equations (3.22) and (3.35), we determine the
coefficient A:

|
\
|

=
3
2
exp ) ( w
C
K
) ( A
2
(3.36)
We also get

) ( w
1
C 2 a
2
1
= and
|
\
|

=
3
2
exp ) ( w
C
K 2
b
2
1
(3.37)
11
and (using (3.18))

|
|
\
|

+ + =
) ( w
) ( w
2 6 ) a 1 ( 2 2
6
1
a
0
(3.38)
Equation (3.21), taking into account (3.38) and (3.26), can be expressed in terms
of F as
F ] ) a a 1 ( 2 [
6
1
F
2
1
F F
2
1
F
2 2 2 3
+ +
|
\
|
+
0 n e
3
K 8
e F K 8
3
2
3
2
= +

(3.39)
where
) 2 a 3 a 3 a 2 (
27
2
54
) a a 1 (
9
n
2 3
3
2
+ +
=
The solution (3.33) is to be substituted in the above equation to get a compatibility
condition between the constants and the various parameters. Using the results of
Appendix A, we get the compatibility conditions
0 ) 2 a 3 a 3 a 2 ( 2 ) 1 a a ( 12 5
2 3 2 3
= + + + + (3.40)
|
|
\
|
|
|
\
|
=
|
|
\
|
|
|
\
|
=
+
+
K 2 3
J
K 2 3
Y
K 2 3
J
K 2 3
Y
C
C
1
1
2
1
(3.41)
0
K 2 3
Y
K 2 3
J
K 2 3
Y
K 2 3
J
1 1
=
|
|
\
|
|
|
\
|
|
|
\
|
|
|
\
|
+ +
(3.42)
We turn now to the determination of the function Y which satisfies Riccati's
equation. Under the substitution

u
u
B
1
Y

= (3.43)
Riccati's equation
2
Y B A Y + = becomes
12
0 u AB u
B
B
u = +
|
\
|

(3.44)
The substitution
y B u = (3.45)
converts (3.44) into
0 y
3
2
exp K y
B
B
2
3
B
B
2
1
y
2
=
|
\
|

+
(
(
\
|

+ (3.46)
taking also into account (3.22). Since
) ( w
C
B
2
= ,we have
w
w
2
B
B
2
3
B
B
2

=
|
\
|

and using (3.32), we obtain, using the various formulas for the derivatives and
recurrence relations of Bessel's functions, that
2 2 2
3
2
exp
9
2
w
w
2 |
\
|

|
\
|

=

Therefore equation (3.46) takes on the form
0 y
3 3
2
exp y
2
2
=
(
(
\
|
\
|

+ (3.47)
where
K 3 K
9
1
2 2 2
= + = (3.48)
Equation (3.47) can be solved by converting it to Bessel's equation, along the lines
of reasoning in transforming (3.27) into (3.31). We thus get the following solution
of equation (3.47):

|
|
\
|
+
|
|
\
|
3
2
3
1
e
K 3 3
Y C
~
e
K 3 3
J C
~
y (3.49)
The same index appears in both expressions of the Bessel's function (3.49) and
(3.33), and is given by (3.34). We then obtain from (3.43), using (3.45) and (3.35)
13

|
|
\
|

=
) ( w
) ( w
) ( y
) ( y
) ( w
C
1
Y
2
(3.50)
Therefore we arrive at the following
Solution I. The solution of equation (3.4)
0 ) ( u ) ( u ) a 1 ( ) ( u a ) ( u ) ( u
3 2
= + + +
is given by (3.5),
Y
b
Y a a ) ( u
1
1 0
+ + = where
|
|
\
|

+ + =
) ( w
) ( w
2 6 ) a 1 ( 2 2
6
1
) ( a
0
,
) ( w
1
C 2 a
2
1
= ,
|
\
|

=
3
2
exp ) ( w
C
K 2
b
2
1
and
|
|
\
|

=
) ( w
) ( w
) ( y
) ( y
) ( w
C
1
Y
2

with ) ( w and ) ( y being expressed in terms of Bessel's functions by (3.33) and
(3.49) respectively. Therefore

) ( w
) ( w
) ( y
) ( y
3
2
exp
K 2
) ( y
) ( y
2 )] a 1 ( 2 2 [
6
1
) ( u
\
|

+ + =
In the above solution we should take into account the condition (A.5), i.e.
6 ) ( w
) ( w
=

. We thus get the solution
1
6 ) ( y
) ( y
3
2
exp K 2
) ( y
) ( y
2 )] a 1 ( 2 [
6
2
) ( u
|
|
\
|
\
|

+ + = (3.51)
where satisfies equation (3.40) and , K satisfy (3.34) and (3.42).
3.II. Case II. For B 2 a
1
= and A 2 b
1
= , we obtain
from (3.14)

A
A
2 3 ) a 1 ( 2 2 a 6
0
+ + + = (3.52)
from (3.10)
14

B
B
2 3 ) a 1 ( 2 2 a 6
0
+ + + = (3.53)
from (3.11)
0 a a 3 a ) a 1 ( 2 B A 8
B
B
B
B
2
0 0
= + + +
+

(3.54)
from (3.13)
0 a a 3 a ) a 1 ( 2 AB 8
A
A
A
A
2
0 0
= + + +
+

(3.55)
from (3.12)
) B A ( 2 3 B A a 12 a a a a a
0
3
0 0 0 0
+ + +
0 a ) a 1 ( AB ) a 1 ( 4 B A 2 2
2
0
= + + + + (3.56)
Equating the two different expressions for
0
a , equations (3.52) and (3.53), we find
that
B
B
A
A
=
and by integration
A s B
2
= (3.57)
where s is real. Equation (3.55) then, using (3.52) and (3.57) becomes
0 A s 8 A m 2 ) A (
2
3
A A
4 2 2 2 2
= + (3.58)
where
] ) a a 1 ( 2 [
6
1
m 2
2 2 2
+ = (3.59)
Equation (3.58) has been solved by a variety of methods in Paper I (Antoniou
[16], Appendices A, B and C). In Appendix C of this paper we list the solution
methods and the corresponding solutions of this equation.
We now turn to the determination of the function Y which satisfies Riccati's
equation. Under the substitution

v
v
B
1
Y

= (3.60)
Riccati's equation
2
Y B A Y + = becomes
15
0 v B A v
B
B
v = +
|
\
|

(3.61)
and because of (3.57),
0 v A s v
A
A
v
2 2
=
|
\
|

(3.62)
From the above equation we obtain A s
v
v
=
and then (see eqs (3.53)-(3.60) of

Paper I, [16])

s
1
) ( Y = (3.63)
Equation (3.56) also gives
A
A
] ) a a 1 ( 2 [
6
1
A
A
A
A
2
3
A
A
2
3
A
A
A
A
3
A
A
2 2
2 3
+ +

+
|
\
|

|
\
|

+

+ + ] ) a a 1 ( 6 [
54
1
A s 8 A A s 24
2 2 2 2 2

0 ) 2 a 3 a 3 a 2 (
27
2
2 3
= + + (3.64)
All the solutions which satisfy (3.58) should be substituted in the above equation.
The resulting expressions will provide compatibility conditions between the
various parameters and constants. According to the results of Appendix B
(equation (B.7)), the compatibility condition (3.604) reads
0 ) 2 a 3 a 3 a 2 ( 2 ) a a 1 ( 6 4
2 3 2 3
= + + + (3.65)
The above equation expresses in terms of the constant a.
Solution II. The solution of equation (3.4)
0 ) ( u ) ( u ) a 1 ( ) ( u a ) ( u ) ( u
3 2
= + + +
is given by (3.5),
Y
b
Y a a ) ( u
1
1 0
+ + = where

|
|
\
|

+ + + =
) ( A
) ( A
2 3 ) a 1 ( 2 2
6
1
) ( a
0
, A s 2 a
2
1
= , A 2 b
1
=
16
and
s
1
Y = with A any solution of the equation (3.58). Therefore
for
s
1
Y = , ) ( A s 2 2
) ( A
) ( A
2
2
] ) a 1 ( 2 2 [
6
1
) ( u

+ + + = (3.66)
for
s
1
Y = , ) ( A s 2 2
) ( A
) ( A
2
2
] ) a 1 ( 2 2 [
6
1
) ( u +

+ + + = (3.67)
The various forms of the function ) ( A are listed in Appendix C and the reader
can easily complete the solutions given above. The parameter satisfies (3.65).
3.III. Case III. For B 2 a
1
= and A 2 b
1
= , we obtain
from (3.14)

A
A
2 3 ) a 1 ( 2 2 a 6
0
+ + = (3.68)
from (3.10)

B
B
2 3 ) a 1 ( 2 2 a 6
0
+ + = (3.69)
from (3.11)
0 a a ) a 1 ( 2 a 3 B A 8
B
B
B
B
0
2
0
= + + +
+

(3.70)
from (3.13)
0 a a ) a 1 ( 2 a 3 B A 8
A
A
A
A
0
2
0
= + + +
+

(3.71)
from (3.12)
) B A ( 2 3 B A a 12 a a a a a
0
3
0 0 0 0
+ +
0 a ) a 1 ( AB ) a 1 ( 4 B A 2 2
2
0
= + + + (3.72)
Equating the two different expressions for
0
conclude that A s B
2
= , which is (3.57). We thus obtain again
s
1
Y = .
Equation (3.71) gives, because of (3.68),
17
0 A s 8 A m 2 ) A (
2
3
A A
4 2 2 2 2
= + (3.73)
where ] ) a a 1 ( 2 [
6
1
m 2
2 2 2
+ = .
Equation (3.72) also gives
A
A
] ) a a 1 ( 2 [
6
1
A
A
A
A
2
3
A
A
2
3
A
A
A
A
3
A
A
2 2
2 3
+ +

+
|
\
|

|
\
|

+

+ + ] ) a a 1 ( 6 [
54
1
A s 8 A A s 24
2 2 2 2 2

0 ) 2 a 3 a 3 a 2 (
27
2
2 3
= + (3.74)
All the solutions which satisfy (3.58) should be substituted in the above equation.
The resulting expressions will provide compatibility conditions between the
various parameters and constants. According to the results of Appendix B
(equation (B.8)), the compatibility condition (3.74) reads
0 ) 2 a 3 a 3 a 2 ( 2 ) a a 1 ( 6 4
2 3 2 3
= + + (3.75)
The above equation expresses in terms of the constant a.
Solution III. The solution of equation (3.4)
0 ) ( u ) ( u ) a 1 ( ) ( u a ) ( u ) ( u
3 2
= + + +
is given by (3.5),
Y
b
Y a a ) ( u
1
1 0
+ + = where

|
|
\
|

+ + =
) ( A
) ( A
2 3 ) a 1 ( 2 2
6
1
) ( a
0
, A s 2 a
2
1
= , A 2 b
1
=
and
s
1
Y = with A any solution of the equation (3.58). Therefore
for
s
1
Y = , A s 2 2
) ( A
) ( A
2
2
] ) a 1 ( 2 2 [
6
1
) ( u +

+ + = (3.76)
for
s
1
Y = , A s 2 2
) ( A
) ( A
2
2
] ) a 1 ( 2 2 [
6
1
) ( u

+ + = (3.77)
18
The various forms of the function ) ( A are listed in Appendix C and the reader
can easily complete the solutions given above. The parameter satisfies (3.75).
3.IV. Case IV. For B 2 a
1
= and A 2 b
1
= , we obtain
from (3.14)

A
A
2 3 ) a 1 ( 2 2 a 6
0
+ + + = (3.78)
from (3.10)

B
B
2 3 ) a 1 ( 2 2 a 6
0
+ + = (3.79)
from (3.11)
0 a a 3 a ) a 1 ( 2 AB 4
B
B
B
B
2
0 0
= + +
+

(3.80)
from (3.13)
0 a a 3 a ) a 1 ( 2 AB 4
A
A
A
A
2
0 0
= + +
+

(3.81)
from (3.12)

3
0
2
0 0 0 0 0
a a ) a 1 ( B A ) a 1 ( 4 B A a 12 a a a a + + + + +
0 ) B A B A ( 2 = (3.82)
Equating the two different expressions of
0
obtain the equation =
|
\
|

+
2
B
B
A
A
3 , which by integration gives

|
\
|

=
3
2
exp K B A (3.83)
Equation (3.82), using (3.79), reads
F e K 8 F ] ) a a 1 ( 2 [
6
1
F
2
1
F F
2
1
F
3
2
2 2 2 3

+ +
|
\
|
+
0 ) 2 a 3 a 3 a 2 (
27
2
54
) a a 1 (
9
e
3
K 8
2 3
3
2
3
2
= +
(3.84)
19
The above equation has to be combined with (3.79), written as
) a 2 a 2 2 (
6
1
3
2
exp K 4 F
2
1
F
2 2 2
+
|
\
|

+ =
Using the results of Appendix A, we get the compatibility conditions (see (A.13)
and (A.10))
0 ) 2 a 3 a 3 a 2 ( 2 ) 1 a a ( 12 5
2 3 2 3
= + + + (3.85)
0
K 2 3
Y
K 2 3
J
K 2 3
Y
K 2 3
J
1 1
=
|
|
\
|
|
|
\
|
|
|
\
|
|
|
\
|
+ +
(3.86)
Using the same reasoning as in Case I, we arrive at the following
Solution IV. The solution of equation (3.4)
0 ) ( u ) ( u ) a 1 ( ) ( u a ) ( u ) ( u
3 2
= + + +
is given by (3.5),
Y
b
Y a a ) ( u
1
1 0
+ + = where
|
|
\
|

+ + + =
) ( w
) ( w
2 6 ) a 1 ( 2 2
6
1
) ( a
0
,
) ( w
1
C 2 a
2
1
= ,
|
\
|

=
3
2
exp ) ( w
C
K 2
b
2
1
and
|
|
\
|

=
) ( w
) ( w
) ( y
) ( y
) ( w
C
1
Y
2

with ) ( w and ) ( y being expressed in terms of Bessel's functions by (3.33) and
(3.49) respectively. Therefore

) ( w
) ( w
) ( y
) ( y
3
2
exp
K 2
) ( y
) ( y
2 )] a 1 ( 2 2 [
6
1
) ( u
\
|

+ + + =
In the above solution we should take into account the condition (A.5), i.e.
6 ) ( w
) ( w
=

. We thus get the solution
1
6 ) ( y
) ( y
3
2
exp K 2
) ( y
) ( y
2 )] a 1 ( 2 [
6
2
) ( u
|
|
\
|
\
|

+ + + = (3.87)
20
where satisfies (3.85) and , K satisfy (3.34) and (3.86).
We list below the derivative of the function ) ( y which appear in Solutions I
and IV. Using the general formulas for the derivatives of Bessel's functions, we
have the following formula

|
|
\
|
+
|
|
\
|
3
1 2
3
1 1
3
e
K 3 3
Y C
~
e
K 3 3
J C
~
e K 3 ) ( y
|
|
\
|
+
|
|
\
|
3
2
3
1
e
K 3 3
Y C
~
e
K 3 3
J C
~
3
1
(3.88)
4. The
|
\
|

G
G
expansion method with variable expansion
coefficients.
We consider that ) ( u is given by an expansion of the form

|
\
|

+ =
G
G
c c ) ( u
1 0
(4.1)
where
0
c and
1
c are dependent coefficients while ) ( G G = is a function
depending on as well. Upon substituting (4.1) into (3.4) and expanding, we
obtain an equation which when arranged in terms of ) ( G , we obtain the
following set of equations:
coefficient of
0
G :
0 c c ) a 1 ( c a c c
3
0
2
0 0 0 0
= + + + (4.2)
coefficient of
1
G
:
G } c c a c c 3 c c ) a 1 ( 2 c {
1 1 1
2
0 1 0 1
+ + +
0 G c G ) c c 2 (
1 1 1
= + + + (4.3)
coefficient of
2
G
:
0 G G c 3 ) G ( } c c c 3 c 2 c ) a 1 ( {
1
2
1 0
2
1 1
2
1
= + (4.4)
21
coefficient of
3
G
:
0 ) G ( c 2 ) G ( c
3
1
3 3
1
= + (4.5)
We first observe that equation (4.2) is essentially equation (4.3) and thus admits
all the solutions found in Section 3.
From equation (4.5), ignoring the trivial solution we conclude that 0 2 c
2
1
= +
and thus 2 c
1
= .
4.I. Case I. For 2 c
1
= , we obtain
from equation (4.4)

0
c 2
3
) a 1 ( 2
G
G
+
=

(4.6)
from equation (4.3)
a )} a 1 ( 2 {
3
c )} a 1 ( 2 { 2 c 3
G
G
0
2
0
+ +
+ + + =

(4.7)
Dividing (4.7) by (4.6) we derive the equation

3
) a 1 ( 2
c 2
a )} a 1 ( 2 {
3
c )} a 1 ( 2 { 2 c 3
G
G
0
0
2
0
+
+
+ +
+ + +
=

(4.8)
In the above equation we substitute any solution of
0
c (which is a solution of
) ( u found in Section 3) and then we solve the third-order ordinary differential
equation to determine the function ) ( G and then the ratio G / G .
4.II. Case II. For 2 c
1
= , we obtain
from equation (4.4)

0
c 2
3
) a 1 ( 2
G
G
+
+ +
=

(4.9)
from equation (4.3)
a )} a 1 ( 2 {
3
c )} a 1 ( 2 { 2 c 3
G
G
0
2
0
+ + +
+ + + =

(4.10)
22
Dividing (4.10) by (4.9) we derive the equation

3
) a 1 ( 2
c 2
a )} a 1 ( 2 {
3
c )} a 1 ( 2 { 2 c 3
G
G
0
0
2
0
+ +
+ + +
+ + +
=

(4.11)
In the above equation we substitute any solution of
0
c (which is a solution of
) ( u found in Section 3) and then we solve the third-order ordinary differential
equation to determine the function ) ( G and then the ratio G / G .
Appendix A. In this Appendix we shall transform the compatibility condition
(3.39) taking into account equation (3.25). Upon substituting
2
F
2
1
F given by
(3.25) into (3.39), we obtain

+ +
3
2
2 2 3
e F K 8 F ] ) a a 1 ( 2 [
6
1
F
2
1
F
0 n ] ) a a 1 ( 2 [
6
e
3
K 4
2 2
3
2
= + +
(A.1)
Multiplying (3.25) by F and rearranging terms, we derive the equation
F ) a 2 a 2 2 (
6
1
3
2
exp KF 4 F F F
2
1
2 2 3
+ +
|
\
|

+ =
Because of the above equation, we obtain from (A.1)

+
3
2
3
2
e
3
K 4
e F K 4 F F F
0 n ] ) a a 1 ( 2 [
6
2 2
= + +
+ (A.2)
Differentiating (3.25) with respect to , we get

=
3
2
e
3
K 8
F F F (A.3)
Combining (A.2) and (A.3) we obtain
23
0 n ] ) a a 1 ( 2 [
6
e
3
K 4
e F K 4
2 2
3
2
3
2
= + +

or
+
|
\
|
+

3
2
e
3
F K 4
0 } ) a a ( 6 ) a 1 ( 4 2 5 ) a a 1 ( 2 12 {
54
2
2 3 3 2
= + + + + (A.4)
Equation (A.4) is true if and only if
0
3
F =
+
and
0 ) a a ( 6 ) a 1 ( 4 2 5 ) a a 1 ( 2 12
2 3 3 2
= + + +
i.e.
0 ) ( w ) ( w 6 = (A.5)
and
0 ) 2 a 3 a 3 a 2 ( 2 ) 1 a a ( 12 5
2 3 2 3
= + + + + (A.6)
respectively, where ) ( w is given by (3.33).
Equation (A.5) should be true for every . Upon expanding (A.8) in a series
around 0 = (this is best facilitated using any of the known Computer Algebra
Systems) we find that (A.5) is true to every order of Taylor's expansion, provided
that
0
K 2 3
Y C
K 2 3
J C
2 1
=
|
|
\
|
+
|
|
\
|

(A.7)
and
0
K 2 3
Y C
K 2 3
J C
1 2 1 1
=
|
|
\
|
+
|
|
\
|
+ +
(A.8)
We thus have
24

|
|
\
|
|
|
\
|
=
|
|
\
|
|
|
\
|
=
+
+
K 2 3
J
K 2 3
Y
K 2 3
J
K 2 3
Y
C
C
1
1
2
1
(A.9)
From the above relations we find that
0
K 2 3
Y
K 2 3
J
K 2 3
Y
K 2 3
J
1 1
=
|
|
\
|
|
|
\
|
|
|
\
|
|
|
\
|
+ +
(A.10)
The above equation provides a relation between , K and a, since is expressed
in terms of a and through (3.34).
Repeating all the above calculations for Case 3.IV, we find that the compatibility
condition (3.84) is equivalent to
+
|
\
|
+

3
2
e
3
F K 4
0 ) a a 1 (
9
4
27
5
) 2 a 3 a 3 a 2 (
27
2
2 3 2 3
= + + + (A.11)
from which we derive the two equations
0 ) ( w ) ( w 6 = (A.12)
and
0 ) 2 a 3 a 3 a 2 ( 2 ) 1 a a ( 12 5
2 3 2 3
= + + + (A.13)
From (A.12) we obtain again (A.9) and (A.10).
Appendix B. In this Appendix we shall simplify the compatibility conditions
(3.64) and (3.74). Introducing the auxiliary function G by

A
A
G

= (B.1)
equation (3.64) transforms into
G ] ) a a 1 ( 2 [
6
1
G
2
1
G G
2
1
G
2 2 2 3
+ +
|
\
|
+
25
+ + ] ) a a 1 ( 6 [
54
1
A s 8 A A s 24
2 2 2 2 2

0 ) 2 a 3 a 3 a 2 (
27
2
2 3
= + + (B.2)
On the other hand, equation (3.58) takes on the form in terms of G:

2 2 2 2
m 2 A s 8 G
2
1
G = (B.3)
with
2
m given by (3.59).
Substituting
2
G
2
1
G given by (B.3) into (B.2), we obtain
G ] ) a a 1 ( 2 [
6
1
G
2
1
G
2 2 3
+ +
+ + ] ) a a 1 ( 6 [
54
1
A A s 24
2 2 2

0 m 2 ) 2 a 3 a 3 a 2 (
27
2
2 2 3
= + + (B.4)
Multiplying (B.3) by G and rearranging terms, we obtain
G m 2 G A s 8 G G G
2
1
2 2 2 3
+ = (B.5)
Equation (B.4), because of (B.5) gives
A A s 24 G A s 8 G G G
2 2 2
+
0 m 2 ) 2 a 3 a 3 a 2 (
27
2
] ) a a 1 ( 6 [
54
1
2 2 3 2 2
= + + + + (B.6)
Differentiating (B.3) with respect to we obtain
) A ( s 8 G G G
2 2
=
and because of the above relation, the first line of equation (B.6) becomes zero and
we thus arrive at
0 m 2 ) 2 a 3 a 3 a 2 (
27
2
] ) a a 1 ( 6 [
54
1
2 2 3 2 2
= + + +
26
or, using (3.59),
0 ) 2 a 3 a 3 a 2 ( 2 ) a a 1 ( 6 4
2 3 2 3
= + + + (B.7)
Repeating all the above calculations for the condition (3.74) of Section 3.III, we
arrive at
0 ) 2 a 3 a 3 a 2 ( 2 ) a a 1 ( 6 4
2 3 2 3
= + + (B.8)
Appendix C.
In this Appendix we list the solution methods and the corresponding solutions of
the equation (3.58):
0 ) ( A s 8 ) ( A m 2 )) ( A (
2
3
) ( A ) ( A
4 2 2 2 2
= + (C.1)
A.I. First Method. We consider an expansion of the form
) ( a a ) ( A
1 0
+ = (C.2)
where ) ( satisfies Jacobi's differential equation

4 2 2
) (
d
d
+ + =
(C.3)
where , and are constant real parameters. Upon substituting (C.2) into
(C.1), taking into account (C.3) and equating to zero the coefficients of the
different powers of , we can determine the coefficients
0
a ,
1
a and the function
) ( . We find that the coefficients
0
a and
1
a are given by the relations
2
2
2
1
s 16
a

= ,
2
2
2
0
s 16
m
a = (C.4)
and the function ) ( by
) m tanh( C
1 ) m tanh( C m
) (
+
+
= (C.5)
where C is an arbitrary constant.
We thus have, using (C.2), the four values (C.4) and equation (C.5), the following
expressions for ) ( A
27

|
|
\
|
+
+
=
) m tanh( C
1 ) m tanh( C m
s 4
m
s 4
) ( A (C.6a)
|
|
\
|
+
+
=
) m tanh( C
1 ) m tanh( C m
s 4
m
s 4
) ( A (C.6b)
|
|
\
|
+
+
=
) m tanh( C
1 ) m tanh( C m
s 4
m
s 4
) ( A (C.6c)
|
|
\
|
+
+
=
) m tanh( C
1 ) m tanh( C m
s 4
m
s 4
) ( A (C.6d)
A.II. Second Method. We consider a ) G / G ( expansion of the form

|
\
|

+ =
G
G
a a ) ( A
1 0
(C.7)
with
0
a and
1
a constants and G a function depending on : ) ( G G = . Upon
substituting (C.7) into (C.1) and equating to zero the coefficients of the different
powers of G, and solving the resulting system, we get
2
2
1
s 16
1
a = ,
2
2
2
0
s 4
m
a = and

|
|
\
|
+ + =
0
1 2
3 2 1
a
a
m 4 exp C C C G (C.8)
Therefore

|
|
\
|
+ +
|
|
\
|

=
0
1 2
0 1 2
0
1 2
1
2
1
a
a
m 4 exp a D D
a
a
m 4 exp a m 4 D
G
G

with
1
D and
2
D constants. We thus have
|
|
|
|
|
\
|
|
|
\
|
+ +
|
|
\
|

+ =
|
\
|

+ =
0
1 2
0 1 2
0
1 2
1
2
1
1 0 1 0
a
a
m 4 exp a D D
a
a
m 4 exp a m 4 D
a a
G
G
a a ) ( A
28
The above expression, gives us the following four expressions (corresponding to
the four combinations of signs for the coefficients
0
a and
1
a ):

|
|
|
|
\
|
+ +

+ =
) m 2 ( exp
s 2
m
D D
) m 2 ( exp
s
m
D
s 4
1
s 2
m
) ( A
1 2
2
1
(C.9a)

|
|
|
|
\
|
+ +
+
=
) m 2 ( exp
s 2
m
D D
) m 2 ( exp
s
m
D
s 4
1
s 2
m
) ( A
1 2
2
1
(C.9b)

|
|
|
|
\
|
+

+ =
) m 2 ( exp
s 2
m
D D
) m 2 ( exp
s
m
D
s 4
1
s 2
m
) ( A
1 2
2
1
(C.9c)

|
|
|
|
\
|
+
+
=
) m 2 ( exp
s 2
m
D D
) m 2 ( exp
s
m
D
s 4
1
s 2
m
) ( A
1 2
2
1
(C.9d)
A.III. Third Method. Using the transformation

) ( w
1
) ( A
2
= (C.10)
equation (C.1) transforms into the Ermakov equation

3 2 2
w s 4 ) ( w m ) ( w

= (C.11)
The solution of Ermakov's equation is given by

=
m 2
1
2
2 2 2 m 2
1 2 2
e
C m 4
s m 16 ) e C C m 2 (
) ( w
We thus obtain, using the above expression and (C.10), that ) ( A is given by

2 2 2 m 2
1 2
m 2
1
2
s m 16 ) e C C m 2 (
e C m 4
) ( A

(C.12a)
29
The above expression can also be written in terms of ) m tanh( as
2 2 2 2
1 2 1 2
2
1
2
)] m ( tanh 1 [ s m 16 )] m ( tanh ) C C m 2 ( ) C C m 2 ( [
)] m ( tanh 1 [ C m 4
) ( A

=
m

(C.12b)
A.IV. Fourth Method. In this case we use the projective Riccati equation
method. We consider the expansion
) ( g b ) ( f a a ) ( A
1 1 0
+ + = (C.13)
where the functions ) ( f and ) ( g satisfy the system
) ( g ) ( f p ) ( f = (C.14)
) ( f r ) ( g p q ) ( g
2
+ = (C.15)

(
(
+
+ = ) ( f
q
r
) ( f r 2 q
p
1
) ( g
2
2
2
(C.16)
The system of equations (C.14) and (C.15) admit five families of solutions
depending on the values of the parameters , p, q and r.
(I) If
2 2
= and 0 q p < then

+ +
=
q p cosh q p sinh r
q
) ( f (C.17)

+ +
+
=
q p cosh q p sinh r
q p sinh q p cosh
p
q p
) ( g (C.18)
with

(
(
+
+ = ) ( f
q
r
) ( f r 2 q
p
1
) ( g
2
2 2 2
2
(C.19)
(II) If
2 2
= and 0 q p > then

+ +
=
q p cos q p sin r
q
) ( f (C.20)
30

+ +
+
=
q p cos q p sin r
q p sin q p cos
p
q p
) ( g (C.21)
with

(
(

+ = ) ( f
q
r
) ( f r 2 q
p
1
) ( g
2
2 2 2
2
(C.22)
(III) If 0 q = , then

+ +
=
2
2
r p
1
) ( f (C.23)
+ +
+
=
2
2
r p
r p
p
1
) ( g (C.24)
with
) ( f
p
r 2
p
) ( f
p
r 2
) ( g
2
2
2
2
(
(
+ = (C.25)
where and are free parameters.
(IV) If 1 p = and
2
r = then
) (
r p
2
r 6
q
) ( f + = (C.26)
) ( 12 q
) ( 12
) ( g
+

= (C.27)
where ) ( satisfies the Weirstrass equation
216
q p
) (
12
q
) ( 4 )) ( (
3 2
3 2
= (C.28)
with solution ) ( ) ( = .
The relation between f and g is given by

q
p
f
p
r 2
g
2
= (C.29)
31
(V) If 1 p = and
25
r
2
= then

) ( 72
q p 5
r 6
q 5
) ( f
2

+ = (C.30)
) ( )] ( 12 q p [
) ( q
) ( g
+

= (C.31)
with ) ( ) ( = and
|
|
\
|
+ =
2
2
2
f
q 25
r 24
f r 2 q
p
1
g (C.32)
Upon substituting (C.13) into (C.1) and equating to zero the coefficients of
i
f
) 4 , 3 , 2 , 1 , 0 i ( = , g and g f
i
) 3 , 2 , 1 i ( = , we find eight solutions, where to every
solution correspond two or three sub-families of solutions.
Family I. This family corresponds to the Solution I
2
0
s 4
m
a = ,
m 2
b
a
1
1
= ,
1 1
b b = ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
where is any root of the equation
2 2
r + = .
From (C.13) and Solution I, we obtain
) ( g b ) ( f
m 2
b
s 4
m
) ( A
1
1
2
+
+ = (C.33)
with
1
2
b s 8 p = ,
1
2
2
b s 2
m
q = with satisfying the equation
2 2
r + = .
Sub-family Ia. Since 0 m 4 q p
2
> = , we consider the choice
2 2
= . The
functions ) ( f and ) ( g in (C.33) are given by (C.17) and (C.18) respectively:

) ( U
1
b s 2
m
) ( f
1
2
2
= and
) ( U
) ( U 2
b s 2
m
) ( g
1
2
2

=
where
32
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + = (C.34)
We thus obtain

) ( U
) ( U m
s 4
1
s 4
m
) ( A
2 2

= (C.35)
2 2 2 2
r + = .
Sub-family Ib. If 1 p = and
2
r = , we have 0 = and then 0 a
1
= . In this
case we have ) ( g b
s 4
m
) ( A
1
2
+ = where ) ( g is given by (C.27).
(Ib.a) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = . Therefore
2
m 4 ) ( 12
) ( 12
) ( g

= and thus
2 2 2
m 4 ) ( 12
) (
s 2
3
s 4
m
) ( A

+ = (C.36)
where ) ( is the Weirstrass function, satisfying the equation

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.37)
(Ib.b) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = . Therefore
2
m 4 ) ( 12
) ( 12
) ( g
+

= and thus
2 2 2
m 4 ) ( 12
) (
s 2
3
s 4
m
) ( A
+

= (C.38)
Sub-family Ic. If 1 p = and
25
r
2
= , satisfies the equation
25
r 24
2
2
= while
) ( f and ) ( g are given by (C.30) and (C.31) respectively.
(Ic.a) For 1 p = , we have
2
1
s 8
1
m 2
a
= ,
2
1
s 8
1
b = and
2
m 4 q = . Therefore
33

) ( 9
m 10
r 3
m 10
) ( f
4 2

+ = and
) ( ] m 4 ) ( 12 [
) ( m 4
) ( g
2
2

=
We thus have the following expression for ) ( A :

] m 4 ) ( 12 [ ) (
) (
s 2
m
r
1
) ( 3
m
s 24
m 5
s 4
m
) ( A
2 2
2 2
2 2

+
|
|
\
|
+ = (C.39)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.40)
(Ic.b) For 1 p = , we have
2
1
s 8
1
m 2
a
= ,
2
1
s 8
1
b = and
2
m 4 q = . Therefore

) ( 9
m 10
r 3
m 10
) ( f
4 2

= and
) ( ] m 4 ) ( 12 [
) ( m 4
) ( g
2
2

=
We thus have the following expression for ) ( A :

] m 4 ) ( 12 [ ) (
) (
s 2
m
r
1
) ( 3
m
s 24
m 5
s 4
m
) ( A
2 2
2 2
2 2

+
|
|
\
|
+ = (C.41)
Note. If
2 2
= and 0 q p > leads to 0 m
2
< . The case 0 q = cannot be
considered either, since it leads to 0 m= and then the coefficient of ) ( f in (C.33)
becomes infinite.
Family II. This family corresponds to the Solution II
2
0
s 4
m
a = ,
m 2
b
a
1
1
= ,
1 1
b b = ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
2 2
r + = .
From (C.13) and Solution II, we obtain
) ( g b ) ( f
m 2
b
s 4
m
) ( A
1
1
2
+
+ = (C.42)
with
1
2
b s 8 p = ,
1
2
2
b s 2
m
q = and satisfies the equation
2 2
r + = .
34
Sub-family IIa. Since 0 m 4 q p
2
> = , we first consider the case
2 2
= .
The functions ) ( f and ) ( g are given by (C.20) and (C.21) respectively:

) ( V
1
b s 2
m
) ( f
1
2
2
= and
) ( V
) ( Z
b s 4
| m | 2
) ( g
1
2
=
where
) | m | 2 ( sin ) | m | 2 ( cos ) ( Z + = (C.43)
) | m | 2 ( cos ) | m | 2 ( sin r ) ( V + + = (C.44)
We thus obtain

) ( V
) ( Z | m | m
s 4
1
s 4
m
) ( A
2 2
+
= (C.45)
2 2 2 2
r = .
Sub-family IIb. If 1 p = and
2
r = then 0 = and thus 0 a
1
= .
In this case we have ) ( g b
s 4
m
) ( A
1
2
+ = where ) ( g is given by (C.27).
(IIb.a) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = . Therefore
2
m 4 ) ( 12
) ( 12
) ( g
+

= and thus
2 2 2
m 4 ) ( 12
) (
s 2
3
s 4
m
) ( A
+

= (C.46)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
= (C.47)
(IIb.b) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = . Therefore
2
m 4 ) ( 12
) ( 12
) ( g

= and thus
35
2 2 2
m 4 ) ( 12
) (
s 2
3
s 4
m
) ( A

+ = (C.48)
Sub-family IIc. If 1 p = and
25
r
2
= then satisfies the equation
25
r 24
2
2
= .
The functions ) ( f and ) ( g are given by (C.30) and (C.31) respectively.
(IIc.a) For 1 p = , we have
2
1
s 8
1
b = ,
2
1
s 8
1
m 2
a
= and
2
m 4 q = . We then
have the following expressions for ) ( f and ) ( g

) ( 9
m 10
r 3
m 10
) ( f
4 2

+ = and
) ( )] ( 12 m 4 [
) ( m 4
) ( g
2
2
+

=
Therefore

)] ( 12 m 4 [ ) (
) (
s 2
m
) ( 3
m
r
1
s 24
m 5
s 4
m
) ( A
2 2
2 2
2 2
+

+
= (C.49)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
= (C.50)
(IIc.b) For 1 p = , we have
2
1
s 8
1
b = ,
2
1
s 8
1
m 2
a
= and
2
m 4 q = . We then
have the following expressions for ) ( f and ) ( g

) ( 9
m 10
r 3
m 10
) ( f
4 2

= and
) ( )] ( 12 m 4 [
) ( m 4
) ( g
2
2
+

=
Therefore

)] ( 12 m 4 [ ) (
) (
s 2
m
) ( 3
m
r
1
s 24
m 5
s 4
m
) ( A
2 2
2 2
2 2
+

+
= (C.51)
Note. If
2 2
= and 0 q p < leads to 0 m
2
< . The case 0 q = cannot be
considered either, since it leads to 0 m= and then the coefficient of ) ( f in (C.42)
becomes infinite.
36
Family III. This family corresponds to the Solution III

2
0
s 4
m
a = ,
1 1
a a = ,
1 1
b b = ,
2
1
2
1
2 2
1
2
) b (
) b ( r ) a ( m 4
= ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
From (C.13) and Solution III, we obtain
) ( g b ) ( f a
s 4
m
) ( A
1 1
2
+ + = (C.52)
where
2
1
2
1
2 2
1
2
) b (
) b ( r ) a ( m 4
= ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
Sub-family IIIa. Since 0 m 4 q p
2
< = for
2 2
= , i.e.
2
1
2
1
2 2
1
2
2 2
) b (
) b ( r ) a ( m 4
= the functions ) ( f and ) ( g are given by (C.17)
and (C.18) respectively:

) ( U
1
s b 2
m
) ( f
2
1
2
= ,
) ( U
) ( U
s 4
1
) ( g
2

=
where
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + = (C.53)
Therefore
) ( U
) ( U
s 4
1
) ( U
1
s b 4
a m 2
s 4
m
) ( A
2 2
1
1
2
2
= (C.54)
where the following relation holds
2
1
2 2 2 2
1
2
b ) r ( a m 4 + = .
Sub-family IIIb. If 1 p = and
2
r = , i.e.
2
1
2
1
2 2
1
2
2
) b (
) b ( r ) a ( m 4
r

= then
0 a
1
= and ) ( g is given by (C.27).
37
(IIIb.a) If 1 p = , then
2
1
s 8
1
b = ,
2
m 4 q = and
2
m 4 ) ( 12
) ( 12
) ( g

= .
Therefore

2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A

+ = (C.55)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.56)
(IIIb.b) If 1 p = , then
2
1
s 8
1
b = ,
2
m 4 q = and
2
m 4 ) ( 12
) ( 12
) ( g
+

= .
Therefore

2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A

= (C.57)
Sub-family IIIc. If 1 p = and
25
r
2
= , i.e.
2
1
2
1
2 2
1
2 2
) b (
) b ( r ) a ( m 4
25
r
=
then ) ( f and ) ( g are given by (C.30) and (C.31) respectively.
(IIIc.a) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = and thus

) ( 9
m 10
r 3
m 10
) ( f
4 2

+ = and
) ( ] m 4 ) ( 12 [
) ( m 4
) ( g
2
2

=
Therefore

] m ) ( 3 [ ) (
) (
s 8
m
) ( 3
m
r
1
3
m a 10
s 4
m
) ( A
2 2
2 2 2
1
2

+
|
|
\
|

+ + = (C.58)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.59)
with
2
1
2 2
1
2
b r 24 a m 100 = .
38
(IIIc.b) For 1 p = , we have
2
1
s 8
1
b = and
2
m 4 q = and thus

) ( 9
m 10
r 3
m 10
) ( f
4 2

= and
) ( ] m 4 ) ( 12 [
) ( m 4
) ( g
2
2

=
Therefore

] m ) ( 3 [ ) (
) (
s 8
m
) ( 3
m
r
1
3
m a 10
s 4
m
) ( A
2 2
2 2 2
1
2

+
|
|
\
|

+ = (C.60)
Note. If
2 2
2
< . The choice 0 q = leads to
0 m= .
Family IV. This family corresponds to the Solution IV
2
0
s 4
m
a = ,
1 1
a a = ,
1 1
b b = ,
2
1
2
1
2 2
1
2
) b (
) b ( r ) a ( m 4
= ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
From (C.13) and Solution IV, we obtain
) ( g b ) ( f a
s 4
m
) ( A
1 1
2
+ + = (C.61)
where
2
1
2
1
2 2
1
2
b
b r a m 4
= ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
Sub-family IVa. Since 0 m 4 q p
2
< = for
2 2
= , i.e.
2
1
2
1
2 2
1
2
2 2
b
b r a m 4
= , then ) ( f and ) ( g are given by (C.17) and (C.18)
respectively:

) ( U
1
b s 2
m
) ( f
1
2
2
= and
) ( U
) ( U
b s 4
1
) ( g
1
2

=
where
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + = (C.62)
We thus have
39
) ( U
) ( U
s 4
1
) ( U
1
b s 2
a m
s 4
m
) ( A
2
1
2
1
2
2
+ = (C.63)
with
2
1
2 2 2 2
1
2
b ) r ( a m 4 + = .
Sub-family IVb. If 1 p = and
2
r = , i.e.
2
1
2
1
2 2
1
2
2
b
b r a m 4
r

= , we have
0 a
1
= .
(IVb.a) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and thus from (C.27) we get
2
m 4 ) ( 12
) ( 12
) ( g

= and then
2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A

= (C.64)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.65)
(IVb.b) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and thus from (C.27) we get
2
m 4 ) ( 12
) ( 12
) ( g
+

= and then
2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A
+

+ = (C.66)
Sub-family IVc. If 1 p = and
25
r
2
= , i.e.
2
1
2
1
2 2
1
2 2
b
b r a m 4
25
r
= then ) ( f and
) ( g are given by (C.30) and (C.31) respectively.
(IVc.a) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and then

) ( 9
m 10
r 3
m 10
) ( f
4 2

+ = and
) ( ] m 4 ) ( 12 [
m 4
) ( g
2
2

=
40
Therefore

] m ) ( 3 [ ) (
) (
s 8
m
) ( 3
m
r
1
3
m a 10
s 4
m
) ( A
2 2
2 2 2
1
2

|
|
\
|

+ + = (C.67)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.68)
with
4 2
2
2
1
s m 800
r 3
a = . This last equation comes from equating the two different
expressions of and using the value of
1
b .
(IVc.b) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and then

) ( 9
m 10
r 3
m 10
) ( f
4 2

= and
) ( ] m 4 ) ( 12 [
m 4
) ( g
2
2

=
Therefore

] m ) ( 3 [ ) (
) (
s 8
m
) ( 3
m
r
1
3
m a 10
s 4
m
) ( A
2 2
2 2 2
1
2

|
|
\
|

+ = (C.69)
Note. The case
2 2
2
< . The choice 0 q =
leads to 0 m= .
Family V. This family corresponds to the Solution V

2
0
s 4
m
a = , 0 a
1
= ,
1 1
b b = ,
1
2
b s 4 p = ,
1
2
2
b s 4
m
q =
From (C.13) and Solution V, we obtain
) ( g b
s 4
m
) ( A
1
2
+ = (C.70)
where
1
2
b s 4 p = ,
1
2
2
b s 4
m
q =
41
Sub-family Va. Since 0 m 4 q p
2
< = , the choice
2 2
= , the function ) ( g
is given by (C.18):
) ( X
) ( X 2
b s 4
m
) ( g
1
2

= where
) | m (| cosh ) | m (| sinh r ) ( X + + = (C.71)
Therefore
) ( X
) ( X
s 2
m
s 4
m
) ( A
2 2

+ = (C.72)
Sub-family Vb. If 1 p = and
2
r = , then ) ( g is given by (C.27).
(Vb.a) For 1 p = , we have
2
1
s 4
1
b = ,
2
m q = and then
2
m ) ( 12
) ( 12
) ( g

= .
Therefore

2 2 2
m ) ( 12
) (
s
3
s 4
m
) ( A

+ = (C.73)

216
m
) (
12
m
) ( 4 )) ( (
6 4
3 2
+ = (C.74)
(Vb.a) For 1 p = , we have
2
1
s 4
1
b = ,
2
m q = and then
2
m ) ( 12
) ( 12
) ( g
+

= .
Therefore
2 2 2
m ) ( 12
) (
s
3
s 4
m
) ( A
+

= (C.75)
Sub-family Vc. If 1 p = and
25
r
2
= then ) ( g is given by (C.31).
(Vc.a) For 1 p = , we get
2
1
s 4
1
b = ,
2
m q = and then
] m ) ( 12 [ ) (
) ( m
) ( g
2
2

= .
Therefore
42

] m ) ( 12 [ ) (
) (
s 4
m
s 4
m
) ( A
2 2
2
2

+ = (C.76)
(Vc.b) For 1 p = , we get
2
1
s 4
1
b = ,
2
m q = and then
] m ) ( 12 [ ) (
) ( m
) ( g
2
2

= . Therefore
] m ) ( 12 [ ) (
) (
s 4
m
s 4
m
) ( A
2 2
2
2

+ = (C.77)
Note. The case
2 2
2
< . The case 0 q = leads
to 0 m= .
Family VI. This family corresponds to the Solution VI

2
0
s 4
m
a = , 0 a
1
= ,
1 1
b b = ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
and r is any root of 0 r
2
= + .
From (C.13) and Solution VI, we obtain
) ( g b
s 4
m
) ( A
1
2
+ = (C.78)
where
1
2
b s 8 p = ,
1
2
2
b s 4
m
q = and r is any root of 0 r
2
= + .
Sub-family VIa. Since 0 m 4 q p
2
< = , the choice
2 2
= , i.e.
0 r
2 2 2
= + and ) ( g is given by (C.18):
) ( U
) ( U
b s 4
1
) ( g
1
2

= ,
where
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + = (C.79)
Therefore
43
) ( U
) ( U
s 4
m
s 4
m
) ( A
2 2

+ = (C.80)
where r satisfies the equation 0 r
2 2 2
= + .
Sub-family VIb. If 1 p = and
2
r = (i.e. r is any real) then ) ( g is given by
(C.27).
(VIb.a) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and thus
2
m 4 ) ( 12
) ( 12
) ( g

= .
Therefore
2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A

+ = (C.81)

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.82)
(VIb.b) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and thus
2
m 4 ) ( 12
) ( 12
) ( g
+

= .
Therefore

2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A
+

= (C.83)
Sub-family VIc. If 1 p = and
25
r
2
= , i.e. 0
25
r
r
2
2
= then ) ( g is given by
(C.31).
(VIc.a) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and then
] m 4 ) ( 12 [ ) (
) ( m 4
) ( g
2
2

= .Therefore
] m ) ( 3 [ ) (
) (
s 8
m
s 4
m
) ( A
2 2
2
2

+ = (C.84)
44
(VIc.b) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and then
] m 4 ) ( 12 [ ) (
) ( m 4
) ( g
2
2

= . Therefore
] m ) ( 3 [ ) (
) (
s 8
m
s 4
m
) ( A
2 2
2
2

+ = (C.85)
Note. The case
2 2
2
to 0 m= .
Family VII. This family corresponds to the Solution VII

2
0
s 4
m
a = , 0 a
1
= ,
1 1
b b = ,
1
2
b s 8 p = ,
1
2
2
b s 2
m
q =
and r is any root of 0 r
2
= + .
From (C.13) and Solution VII, we obtain
) ( g b
s 4
m
) ( A
1
2
+ = (C.86)
where
1
2
b s 8 p = ,
1
2
2
b s 4
m
q = and r is any root of 0 r
2
= + .
Sub-family VIIa. Since 0 m 4 q p
2
< = , the choice
2 2
= , i.e.
0 r
2 2 2
= + , then ) ( g is given by (A.18):
) ( U
) ( U
b s 4
1
) ( g
1
2

= ,
where
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + =
Therefore

) ( U
) ( U
s 4
1
s 4
m
) ( A
2 2

= (C.87)
where r satisfies the equation 0 r
2 2 2
= + .
45
Sub-family VIIb. If 1 p = and
2
r = (i.e. r any real) then ) ( g is given by
(C.27).
(VIIb.a) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and thus
2
m 4 ) ( 12
) ( 12
) ( g

= . Therefore
2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A

= (C.88)
where ) ( is Weirstrass function satisfying the equation

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.89)
(VIIb.b) For 1 p = , we have
2
1
s 8
1
b = ,
2
m 4 q = and thus
2
m 4 ) ( 12
) ( 12
) ( g
+

= .
Therefore
2 2 2
m ) ( 3
) (
s 8
3
s 4
m
) ( A
+

+ = (C.90)
Sub-family VIIc. If 1 p = and
25
r
2
= , i.e. 0
25
r
r
2
2
= then ) ( g is given by
(C.31).
(VIIc.a) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and then
] m 4 ) ( 12 [ ) (
) ( m 4
) ( g
2
2

= .Therefore
] m ) ( 3 [ ) (
) (
s 8
m
s 4
m
) ( A
2 2
2
2

= (C.91)
(VIIc.b) For 1 p = , we get
2
1
s 8
1
b = ,
2
m 4 q = and then
46
] m 4 ) ( 12 [ ) (
) ( m 4
) ( g
2
2

= . Therefore
] m ) ( 3 [ ) (
) (
s 8
m
s 4
m
) ( A
2 2
2
2

= (C.92)
Note. The case
2 2
2
to 0 m= .
Family VIII. This family corresponds to the Solution VIII
0 a
0
= ,
m s 8
p
a
2
1
= , 0 b
1
= , p p = ,
p
m 4
q
2
=
and is any root of
2 2
r + = .
From (C.13) and Solution VIII, we obtain
) ( f
m s 8
p
) ( A
2

= (C.93)
where p p = ,
p
m 4
q
2
= and is any root of the equation
2 2
r + = .
Sub-family VIIIa. Since 0 m 4 q p
2
< = , the choice
2 2
= , i.e.
2 2 2 2
r + = , then ) ( f is given by (C.17):
) ( U
1
p
m 4
) ( f
2
= where
) | m | 2 ( cosh ) | m | 2 ( sinh r ) ( U + + = (C.94)
Therefore
) ( U
1
s 2
m
) ( A
2
= (C.95)
where satisfies the equation
2 2 2 2
r + = .
Sub-family VIIIb. If 1 p = and
25
r
2
= , leads to
2 2
r
25
24
= and then ) ( f is
given by (C.30).
47
(VIIIb.a) For 1 p = , we have
2
m 4 q = and thus
) ( 9
m 10
r 3
m 10
) ( f
4 2

+ = .

27
m 8
) (
3
m 4
) ( 4 )) ( (
6 4
3 2
+ = (C.96)
Therefore
|
|
\
|

+
=
) ( 3
m
r
1
s 12
m 5
) ( A
2
2
(C.97)
and is any root of the equation
25
r 24
2
2
= .
(VIIIb.b) For 1 p = , we have
2
m 4 q = and thus
) ( 9
m 10
r 3
m 10
) ( f
4 2

= .
Therefore
|
|
\
|

=
) ( 3
m
r
1
s 12
m 5
) ( A
2
2
(C.98)
Note. The case
2 2
2
to 0 m= . The case
2
r = cannot be considered since in this case 0 = and then
0 a
1
= (i.e. 0 ) ( A = ).
A.V. Fifth Method. We consider now an expression for ) ( A of the form

+ +
+ +
=
m
1 0
m
1
m
1 0
m
1
e d c e c
e b a e a
) ( A (C.99)
and substitute back into (C.1). We then derive the following set of solutions

+
=
m
2
m
1
m
e C e C
e
) ( A (C.100)

=
m m
1
m
e s 2 e C m
e m
) ( A (C.101)
48

+
=
m m
1
m
e s 2 e C m
e m
) ( A (C.102)

2 m
2
2 m
1
m
2
m
1
2
s 4 e C s 4 e C s m 2
s m 2 e C s m 2 e C m
) ( A
+ +
+ +
=

(C.103)
A.VI. Sixth Method. We consider an expansion of the form
) ( a a ) ( A
1 0
+ = (C.104)
where ) ( satisfies Jacobi's differential equation

4
4
3
3
2
2 1 0
n n n n n ) (
d
d
+ + + + =
(C.105)
Upon substituting (C.104) into (C.1), taking into account (C.105) and equating to
zero the coefficients of the different powers of to zero, we find eleven
solutions. For the first three solutions, we use the following Lemma:
Lemma: If (C.105) is expressed as
) D q p ( ) D q p ( ) n ( r ) (
d
d
+ + =
(C.106)
then its solution is given by

D q 4 e ) p n ( 4 e
D q n 4 e ) p n p D q ( 4 e n
2 K r K r 2
2 K r 2 2 K r 2
+ + +
+ +
=

(C.107)
where
D q ) p n ( K
2 2
+ = (C.108)
For the Solution 1, we have
|
|
\
|
+ = + + + +
2
1
0 2
1
2 4
4
3
3
2
2 1 0
a
a
a s 16 n n n n n
|
|
\
|

|
|
\
|
+

2
1
2 2
1
2
3
2
1 0
2
1
2 2
1
2
3
2
1 0
a s 32
D
a s 32
n s a a 32
a s 32
D
a s 32
n s a a 32

where

2
1
2
3
2
1 0
) a s m 16 ( ) n s a a 64 ( D
49
Equation (C.105) then gives by integration the relation (C.107), where

2
1
2
3
2
1 0
a s 32
n s a a 32
p

= and
2
1
2
a s 32
1
q = (C.109)
+
= + + + +
2
2
2 2
0
2
2
3
2
0
2
4
4
3
3
2
2 1 0
) n m 4 a s 96 (
n a s 144
n n n n n
|
|
\
|
+
+
2
3
2
2 2
0
2
n 3
n m 4 a s 96

|
|
\
|
+
+
+

3
2
0
2
2
2 2
0
2
3
2
0
2
2
2
2
2 2
0
2
n a s 288
D ) n m 4 a s 96 (
n a s 288
) n m 4 ( ) n m 4 a s 96 (

|
|
\
|
+
+

3
2
0
2
2
2 2
0
2
3
2
0
2
2
2
2
2 2
0
2
n a s 288
D ) n m 4 a s 96 (
n a s 288
) n m 4 ( ) n m 4 a s 96 (

where

2
0
2
2
2 2
0
2
) a s m 48 ( ) n m 4 a s 96 ( D +
(C.105) then gives by integration the relation (A.107), where

3
2
0
2
2
2
2
2 2
0
2
n a s 288
) n m 4 ( ) n m 4 a s 96 (
p
+
= ,
3
2
0
2
2
2 2
0
2
n a s 288
n m 4 a s 96
q
+
= (C.110)

= + + + +
2
1
2
2
2 2
0
2 2
0
2
4
4
3
3
2
2 1 0
n
) n m 4 a s 32 ( a s 16
n n n n n
|
|
\
|
+
+
2
2
0
2
2
2
1
a s 32 n m 4
n

|
|
\
|

+

) n m 4 a s 32 ( a s 96
D n
) n m 4 a s 32 ( a s 96
) m 4 n ( n
2
2 2
0
2 2
0
2
1
2
2 2
0
2 2
0
2
2
2 1

50
|
|
\
|

) n m 4 a s 32 ( a s 96
D n
) n m 4 a s 32 ( a s 96
) m 4 n ( n
2
2 2
0
2 2
0
2
1
2
2 2
0
2 2
0
2
2
2 1

where

2
0
2
2
2 2
0
2
) a s m 48 ( ) n m 4 a s 96 ( D +
(C.105) then gives by integration the relation (A.107), where

) n m 4 a s 32 ( a s 96
) m 4 n ( n
p
2
2 2
0
2 2
0
2
2
2 1

= ,
) n m 4 a s 32 ( a s 96
n
q
2
2 2
0
2 2
0
2
1

= (C.111)
The other eight solutions are not considered here, since they lead to very
complicated expressions. They will be published elsewhere in electronic form.
A.VII. Seventh Method. We consider that ) ( A has the form

|
\
|

+ =
G
G
a a ) ( A
1 0
(C.112)
where
0
a and
1
a are dependent quantities, ) ( a a
0 0
= and ) ( a a
1 1
= , while
) ( G G = . Upon substituting (C.112) into (C.1) and equating to zero the
coefficients of the different powers of G, we obtain a system of ordinary
differential equations, from which we find
s
1
a
1
= and that ) ( a
0
satisfies the
equation
0 a s 8 a m 2 ) a (
2
3
a a
4
0
2 2
0
2 2
0 0 0
= + (C.113)
We also find that the ratio G / G is given by
) s , m ( F
G
G
=

(C.114)
where

1 1
0
0
0
1 0
0 2
0
2 2
1
0
2
0
0
a 3
D
a
a
3
a
a
D
a a
a
a s 32 m 4
a
a
9
a
a
3
) s , m ( F
|
|
\
|
+
|
|
\
|
|
|
\
|
= (C.115)
51
with
2
2
0 1
2
0
2
0
0
2
s 4
m 3
a a 36 a 75
a
a
s 16
9
D +
|
|
\
|
= (C.116)
Equation (C.113) is essentially equation (C.1) and admits all the solutions
equation (C.1) admits. When a solution is substituted in (C.115), we can integrate
in principle equation (C.114). We then can evaluate the function ) ( G and then
the ratio G / G . The function ) ( A is determined from (C.112).
References
[1] S. Abbasbandy: Soliton solutions for the Fitzhugh-Nagumo equation with
the homotopy analysis method.
Appl. Math. Model. 32 (2008) 2706-2714
[2] P. C. Abbott, E. J. Parkes and B. R. Duffy:The Jacobi elliptic-function
method for finding periodic-wave solutions to nonlinear evolution
equations. Available online at
http//physics.uwa.edu.au/pub/Mathematica/Solitons
[3] M. A. Abdou: Adomian decomposition method for solving the telegraph
equation in charged particle transport.
J. Quant. Spectro. Rad. Trans. 95 (2005) 407-414
[4] M. A. Abdou: Exact solutions for nonlinear evolution equations via
the extended projective Riccati equation expansion method.
Elec. J. Th. Phys. 4 (2007) 17-30.
[5] M. A. Abdou: The extended F-expansion method and its applications
for a class of nonlinear evolution equations.
Chaos, Solitons and Fractals 31 (2007) 95-104.
[6] M. A. Abdou: On the variational iteration method.
Phys. Lett. A 366 (2007) 61-68
[7] M. A. Abdou:A generalized auxiliary equation method and its
applications. J. Nonl. Dyn. 52 (2008) 95-102
52
[8] M. A. Abdou: An improved generalized F-expansion method and its
applications. J. Comp. Appl. Math. 214 (2008) 202-208
[9] M. A. Abdou: Generalized solitary and periodic solutions for nonlinear
partial differential equations by the exp-function method.
J. Nonl. Dyn. 52 (2008) 1-9
[10] M. A. Abdou and A. A. Soliman: New applications of Variational
Iteration Method. Physica D 211 (2005) 1-8
[11] M. A. Abdou and A. Elhanbaly: Construction of periodic and solitary
wave solutions by the extended Jacobi elliptic function expansion method".
Comm. Nonl. Sci. Num. Sim. 12 (2007) 1229-1241
[12] M. J. Ablowitz, D. J. Kaup, A. C. Newell and H. Segur: The Inverse
Scattering Transform. Fourier Analysis for Nonlinear Problems.
Stud. Appl. Math. 53 (1974) 249-315
[13] M. J. Ablowitz and P. A. Clarkson: Solitons, Nonlinear Evolution
Equations and Inverse Scattering Transform.
Cambridge University Press, Cambridge, 1991.
[14] M. J. Ablowitz and H. Segur: Solitons and the Inverse Scattering
Transform. SIAM, 1981.
[15] G. Adomian: Nonlinear Stochastic Operator Equations.
Academic Press, San Diego, (1986).
[16] S. Antoniou: The Riccati equation method with variable expansion
coefficients. I. Solving the Burgers equation. submitted for publication
[17] S. Antoniou: The Riccati equation method with variable expansion
coefficients. II. Solving the KdV equation. submitted for publication
[18] I. Aslan: Application of the exp-function method to nonlinear lattice
differential equations for multi-wave and rational solutions.
Math. Meth. Appl. Sci. 60 (2011) 1707-1710.
[19] D. Baldwin, . Gkta, W. Hereman, L. Hong, R.S. Martino and
J.C. Miller: Symbolic computation of exact solutions expressible in
53
hyperbolic and elliptic functions for nonlinear PDEs.
J. Symb. Comp. 37 (2004) 669-705
[20] A. Bekir and A. Boz: Exact Solutions for Nonlinear Evolution Equations
using Exp-Function Method. Phys. Lett. A 372 (2008) 1619-1625.
[21] E. D. Belokolos, A. I. Bobenko, V. Z. Enolskii, A. R. Its and V.
Matveev: "Algebro-Geometric Approach to Nonlinear Integral Equations"
Springer-Verlag 1994
[22] G. Bluman and S. Kumei: Symmetries and Differential Equations.
Springer-Verlag 1989
[23] A. Borhanifar and A. Z. Moghanlu: Application of the ) G / G (
expansion method for the Zhiber-Sabat equation and other related
equations. Math. Comp. Mod. 54 (2011) 2109-2116.
[24] H. T. Chen and H. Q. Zhang: Improved Jacobian elliptic function
method and its applications.
Chaos, Solitons and Fractals 15 (2003) 585-591
[25] Y. Chen and Q. Wang: Extended Jacobi elliptic function rational
expansion method and abundant families of Jacobi elliptic function
solutions to (1+1)-dimensional dispersive long wave equation.
[26] C. Dengyuan and G. Yuan: "Cole-Hopf quotient and exact solutions of
the generalized Fitzhugh-Nagumo equations".
Acta Math. Scientia 19 (1999) 7-14
[27] A. E. Ebaid: Generalization of Hes Exp-Function Method and New
Exact Solutions for Burgers Equation. Z. Naturforsch. 64a (2009) 604608
[28] S. A. El-Wakil, E. M. Abulwafa, A. Elhanbaly and M. A. Abdou:The
extended homogeneous balance method and its applications.
[29] S. A. El-Wakil and M. A. Abdou: New exact travelling wave solutions
using modified extended tanh-function method.
54
[30] S. A. El-Wakil, M. A. Abdou and A. Hendi: New periodic wave
solutions via exp-function method. Phys. Lett. A 372 (2008) 830-840
[31] S. A. El-Wakil and M. A. Abdou: "New applications of the homotopy
analysis method". Z. Naturforsch. A (2008)
[32] P. G. Estevez: The direct method and the singular manifold method for
the Fitzhugh-Nagumo equation. Phys. Lett. A 171 (1992) 259-261
[33] E. Fan: Two new applications of the homogeneous balance method
Phys. Lett. A 265 (2000) 353-357
[34] E. Fan: Extended tanh-function method and its applications to nonlinear
equations . Phys. Lett. A 277 (2000) 212-218.
[35] E. Fan and Y. C. Hon: Applications of extended tanh-method to
special types of nonlinear equations.
Appl. Math. Comp. 141 (2003) 351-358.
[36] E. Fan and H. Zhang: Applications of the Jacobi elliptic function method
to special-type nonlinear equations. Phys. Lett. A 305 (2002) 383-392.
[37] J. Feng, W. Li and Q. Wan: Using ) G / G ( expansion method to seek
traveling wave solution of Kolmogorov-Petrovskii-Piskunov equation
Appl. Math. Comp. 217 (2011) 5860-5865.
[38] Z. S. Feng: "The first integral method to study the Burgers-Korteweg de
Vries equation". J. Phys. A: Math. Gen. A 302 (2002) 343-349
[39] R. FitzHugh: "Mathematical models of threshold phenomena in the nerve
membrane". Bull. Math. Bioph. 17 (1955) 257278
[40] R. FitzHugh: " Impulses and physiological states in theoretical models of
nerve membrane". Bioph. J. 1 (1961) 445466
[41] R. FitzHugh: " Mathematical models of excitation and propagation in
nerve". (in H.P. Schwan (ed). "Biological Engineering" McGrawHill,
N.Y. 1969)
[42] P. Gray and S. Scott: Chemical Oscillations and Instabilities
55
Clarendon Press, Oxford, 1990.
[43] P. Griffiths and W. E. Sciesser: Traveling Wave Analysis of Partial
Differential Equations. Academic Press 2012.
[44] J. H. He: A variational iteration method-a kind of nonlinear analytical
technique: Some examples. Int. J. Nonlin. Mech. 34 (1999) 699708
[45] J. H. He and M. A. Abdou: New periodic solutions for nonlinear
evolution equations using exp-function method.
[46] J. H. He and X. H. Wu: Exp-function method for nonlinear wave
equations. Chaos, Solitons and Fractals 30 (2006) 700-708.
[47] W. Hereman and W. Malfliet: The tanh method: A Tool to Solve
Nonlinear Partial Differential Equations with Symbolic Software.
Available online, Colorado School of Mines.
[48] R. Hirota: The Direct Method in Soliton Theory
Cambridge University Press 2004.
[49] R. Hirota: Exact Solution of the KdV Equation for Multiple Collisions of
Solitons. Phys. Rev. Lett. 27 (1971) 1192-1194.
[50] P. E. Hydon: Symmetry Methods for Differential Equations
Cambridge University Press 2000
[51] M. Inc and M. Ergt: Periodic wave solutions for the generalized
shallow water wave equation by the improved Jacobi elliptic function
method. Appl. Math. E-Notes 5 (2005) 89-96
[52] A. Jabbari, H. Kheiri and A. Bekir : Exact solutions of the coupled
Higgs equation and the Maccari system using Hes semi-inverse method
and ) G / G ( expansion method.
Comp. Math. Appl. 62 (2011) 2177-2186.
[53] A. J. M. Jawad, M. D. Petkovich and A. Biswas : Modified simple
equation method for nonlinear evolution equations.
Appl. Math. Comp. 217 (2010) 869877
56
[54] Y. Keskin and G. Oturanc: Reduced Differential Transform Method
for Partial Differential Equations".
Int. J. Nonl. Sci. Num. Sim. 10 (2009) 741-749
[55] N. A. Kudryashov: Exact Solutions of the Generalized Kuramoto-
Sivashinsky Equation. Phys. Lett. A 147 (1990) 287-291.
[56] N. A. Kudryashov: Simplest equation method to look for exact solutions
of nonlinear differential equations. arXiv:nlin/0406007v1, 4 Jun 2004
[57] N. A. Kudryashov: Nonlinear differential equations with exact solutions
expressed via the Weirstrass function. arXiv:nlin/0312035v1,16 Dec 2003
[58] H. Li and Y. Guo: "New exact solutions to the FitzhughNagumo
equation". Appl. Math. Comp. 180 (2006) 524528.
[59] S. Liao: Homotopy Analysis Method in Nonlinear Differential
Equations. Springer 2012
[60] S. K. Liu, Z. Fu, S. Liu and Q. Zhao : Jacobi elliptic function
expansion method and periodic wave solutions of nonlinear wave
equations. Phy. Lett. A 289 (2001) 69-74.
[61] S. K. Liu, Z. Fu, S. Liu and Q. Zhao : Expansion about the Jacobi
Elliptic Function and its applications to Nonlinear Wave Equations.
Acta Phys. Sinica 50 (2001) 2068-2072.
[62] D. Lu and Q. Shi: New Jacobi elliptic functions solutions for the
combined KdV-mKdV Equation. Int. J. Nonlin. Sci. 10 (2010) 320-325
[63] W. Malfliet: Solitary wave solutions of nonlinear wave equations.
Am. J. Phys. 60 (1992) 650-654.
[64] W. Malfliet: The tanh method: a tool for solving certain classes of
nonlinear evolution and wave equations
J. Comp. Appl. Math. 164-165 (2004) 529-541
[65] W. Malfliet and W. Hereman: The tanh method: I. Exact solutions of
nonlinear evolution and wave equations.
Physica Scripta 54 (1996) 563-568
57
[66] W. Malfliet and W. Hereman: The tanh method: II. Perturbation
technique for conservative systems. Physica Scripta 54 (1996) 569-575.
[67] J. Murray: Mathematical Biology. Springer-Verlag, Berlin, 1989.
[68] J. Nagumo, S. Arimoto and S. Yoshizawa: "An active pulse transmission
line simulating nerve axon". Proc. IRE. 50 (1962) 20612070.
[69] H. Naher, F. Abdullah and M. A. Akbar: The exp-function method for
new exact solutions of the nonlinear partial differential equations.
Int. J. Phys. Sci. 6 (2011) 6706- 6716.
[70] H. Naher, F. A. Abdullah and M. A. Akbar: New travelling wave
solutions of the higher dimensional nonlinear partial differential
equation by the exp-function method. J. Appl. Math. (2012)
[71] H. Naher, F. A. Abdullah and M. A. Akbar:The ) G / G ( expansion
method for abundant traveling wave solutions of Caudrey-Dodd-Gibbon
equation. Math. Prob. Engin. (2011)
[72] S. P. Novikov, S. V. Manakov, L. P. Pitaevskii and V. E. Zakharov:
Theory of Solitons: The Inverse Scattering Method. Plenum, NY 1984
[73] M. C. Nucci and P.A. Clarkson: "The nonclassical method is more
general than the direct method for symmetry reductions: an example of the
Fitzhugh-Nagumo equation". Phys. Lett. A 164 (1992) 4956.
[74] P. J. Olver: Applications of Lie Groups to Differential Equations
Graduate Texts in Mathematics, vol.107, Springer Verlag, N.Y. 1993
[75] L. V. Ovsiannikov: Group Analysis of Differential Equations
Academic Press, New York 1982
[76] T. Ozis and I. Aslan: Application of the ) G / G ( expansion method
to Kawahara type equations using symbolic computation.
Appl. Math. Comp. 216 (2010) 2360-2365.
[77] E. J. Parkes and B. R. Duffy: An automated tanh-function method for
finding solitary wave solutions to nonlinear evolution equations.
Comp. Phys. Comm. 98 (1996) 288-300
58
[78] E. J. Parkes, B. R. Duffy and P. C. Abbott: The Jacobi elliptic function
method for finding periodic-wave solutions to nonlinear evolution
equations. Phys. Lett. A 295 (2002) 280-286
[79] E. J. Parkes, E. J. Zhu, B. R. Duffy and H. C. Huang: Sech-polynomial
traveling solitary-wave solutions of odd-order generalized KdV equations
Phys. Lett. A 248 (1998) 219-224
[80] K. R. Raslan: The first integral method for solving some important
nonlinear partial differential equations. Nonl. Dyn. (2007)
[81] J. Rauch and J. Smoller: "Qualitative Theory of the FitzHugh-Nagumo
Equations". Adv. Math. 27 (1978) 12-44
[82] C. Rogers and W. F. Shadwick: Bcklund Transformations.
Academic Press, New York, 1982.
[83] W. Rui, S. Xie, Y. Long and B. He: "Integral Bifurcation Method and its
Applications for solving the modified Equal Width Wave equation and its
variants". Rostock Math. Kolloq. 62 (2007) 87-106
[84] A. H. Salas, and C. A. Gomez : Application of the Cole-Hopf
transformation for finding exact solutions to several forms of the
seventh-order KdV equation. Math. Prob. Eng. (2010)
[85] A. A. Soliman and H. A. Abdou: New exact solutions of nonlinear
variants of the RLW, the phi-four and Boussinesq equations based on
modified extended direct algebraic method.
Int. J. Nonl. Sci. 7 (2009) 274-282.
[86] H. Stephani: Differential Equations: Their Solutions Using Symmetries.
Cambridge University Press, 1989
[87] N. Taghizadeh, M. Akbari and A. Ghelichzadeh: Exact solution of
Burgers equations by homotopy perturbation method and reduced
differential transformation method.
Austr. J. Basic Appl. Sci. 5 (2011) 580-589
[88] N. K. Vitanov: Application of simplest equations of Bernoulli and Riccati
59
kind for obtaining exact traveling-wave solutions for a class of PDEs
with polynomial nonlinearity.
Comm. Nonlin. Sci. Num. Sim. 15 (2010) 20502060
[89] M. L. Wang, Y. B. Zhou and Z. B. Li: Application of a homogeneous
balance method to exact solutions of nonlinear equations in
Mathematical Physics. Phys. Lett. A 216 (1996) 67-75.
[90] M. L. Wang and X. Z. Li: Applications of F-Expansion to periodic wave
solutions for a new Hamiltonian amplitude equation.
[91] M. Wang, X. Li and J. Zhang: The ) G / G ( expansion method and
travelling wave solutions of nonlinear evolution equations in
Mathematical Physics. Phys. Lett. A 372 (2008) 417-421.
[92] A.M. Wazwaz: A reliable modification of Adomians decomposition
Method. Appl. Math.Comp. 92 (1998) 17.
[93] A. M. Wazwaz: The tanh-coth method for solitons and kink solutions
for nonlinear parabolic equations.
Appl. Math. Comp. 188 (2007) 1467-1475.
[94] A. M. Wazwaz: Partial Differential Equations and Solitary Waves
Theory. Springer-Verlag Berlin Heidelberg 2009
[95] J. Weiss, M. Tabor and G. Carnevale: The Painleve Property for Partial
Differential Equations. J. Math. Phys. 24 (1982) 522-526.
[96] J. Weiss, M. Tabor and G. Carnevale: The Painleve Property .
J. Math. Phys. 24 (1983) 1405-
[97] G. Whitham: Linear and Nonlinear Waves. Wiley, NY 1974
[98] K. Yahya, J. Biafar, H. Azari and P. R. Fard: Homotopy Perturbation
Method for Image Restoration and Denoising. Available online.
[99] O. Yu. Yefimova: The modified simplest equation method to look for
exact solutions of nonlinear partial differential equations.
arXiv:1011.4606v1 [nlin.SI] 20 Nov 2010
60
[100] E. M. E. Zayed: Traveling wave solutions for higher dimensional
nonlinear evolution equations using the ) G / G ( expansion method.
J. Appl. Math. Inf. 28 (2010) 383- 395
[101] E. M. E. Zayed: A note on the modified simple equation method applied
to Sharma-Tasso-Olver equation.
Appl. Math. Comp. 218 (2011) 3962-3964
[102] E. M. E. Zayed and K. A. Gepreel: The (G/G)-expansion method for
finding traveling wave solutions of nonlinear partial differential equations
in mathematical physics. J. Math. Phys. 50 (2009) 013502
[103] X. L. Zhang and H. Q. Zhang: A new generalized Riccati equation
rational expansion method to a class of nonlinear evolution equations with
nonlinear terms of any order. Appl. Math. Comp.186 (2007) 705-714

The Riccati Equation Method With Variable Expansion Coefficients. IV. Solving The Fitzhugh-Nagumo Equation.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Riccati Equation Method With Variable Expansion Coefficients. IV. Solving The Fitzhugh-Nagumo Equation.

Uploaded by

Copyright:

Available Formats

The Riccati equation method

with variable expansion coefficients.

and then (see eqs (3.53)-(3.60) of

You might also like