THE Q-HOMOTOPY ANALYSIS METHOD (Q-HAM) - IJAMM
THE Q-HOMOTOPY ANALYSIS METHOD (Q-HAM) - IJAMM
THE Q-HOMOTOPY ANALYSIS METHOD (Q-HAM) - IJAMM
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
ABSTRACT<br />
<strong>THE</strong> Q-<strong>HOMOTOPY</strong> <strong>ANALYSIS</strong> <strong>METHOD</strong> (Q-<strong>HAM</strong>)<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil 1 and S. N. Huseen 2<br />
1 Cairo University, Faculty of Engineering,<br />
Engineering Mathematics Department, Giza, Egypt<br />
Email: magdyeltawil@yahoo.com<br />
2 Thi-Qar University, Faculty of Computer Science and Mathematics,<br />
Mathematics Department, Thi-Qar, Iraq<br />
Email: shn_n2002@yahoo.com<br />
Received 30 November 2011; accepted 27 March 2012<br />
In this paper, a more general method of homotopy analysis method (<strong>HAM</strong>) is introduced to<br />
solve non-linear differential equations, it is called (q-<strong>HAM</strong>). The interval of convergence of<br />
<strong>HAM</strong>, if exists, is increased when using q-<strong>HAM</strong>. The analysis shows that the series solution<br />
in the case of q-<strong>HAM</strong> is more likely to converge than that on <strong>HAM</strong>. The new method is<br />
applied to some nonlinear differential equations to illustrate the method of analysis.<br />
Keywords: Homotopy Analysis Method (<strong>HAM</strong>), Partial Differential Equations<br />
1 INTRODUCTION<br />
The standard homotopy analysis method (<strong>HAM</strong>) is an analytic method that provides series<br />
solutions for nonlinear partial differential equations and has been firstly proposed by (Liao<br />
1992). In recent years, this method has been successfully employed to solve many types of<br />
linear and non-linear differential equations in various fields of engineering and science (Liao<br />
1995,1996,1997,2002,2003,2004,2009 ;Wang et al.2007;Bataineh et al.2009;Van Gorder and<br />
Vairavelu 2009). Different from all perturbation techniques ( the techniques are strongly<br />
dependent upon small / large physical parameters (Nayfeh 1981,1985,2000;Lagerstrom<br />
1988;Murdock 1991;Hinch 1991)) and non perturbation techniques, such as the Lyapunov<br />
artificial small parameter method (Lyapunov 1992) , the -expansion method (Karmishin et<br />
al. 1990; Awrejcewicz et al. 1998) , Adomians decomposition method (Rach 1984; Adomain<br />
and Adomain GE 1984,Adomain 1991) and so on which are formally independent of<br />
small/large physical parameters but they can not ensure the convergence of solution series,<br />
the homotopy analysis method has the following advantages:<br />
(i) It is valid even if a given nonlinear problem does not contain any small/large<br />
parameters.<br />
(ii) It can provide us with a convenient way to adjust and control the convergence<br />
region and rate of approximation series.<br />
(iii) It can be employed to efficiently approximate a nonlinear problem by choosing<br />
different sets of base functions.
52<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
In recent years more and more researchers have been successfully applying (<strong>HAM</strong>) to various<br />
nonlinear problems in science and engineering ,such as, the nonlinear water waves (Tao et al.<br />
2007, groundwater flows (Song and Tao 2007) , time-dependent Emden-Fowler type<br />
equations (Bataineh et al. 2007) , solving high order nonlinear differential equations (Hany N<br />
Hassan and El-Tawil 2010), solving two points nonlinear boundary value problems (Hany N<br />
Hassan and El-Tawil 2010), time-fractional partial differential equations (O Abdulaziz et al.<br />
2008) ,Klein- Gordon equation (Alomari et al. 2008 ,Goursat problems(Syyed Ali<br />
Kazemipour and Ahmad Neyrameh 2010) ,one group neutron diffusion equation(Cavdar<br />
2010) ,algebraic equations (Chin Fugn Yuen et al. 2010 ,integral and integro-differential<br />
equations (Hossein Zadeh et al. 2010), singular higher-order boundary value problems(A<br />
Sami Bataineh et al. 2011).<br />
2 BASIC IDEA OF Q-<strong>HOMOTOPY</strong> <strong>ANALYSIS</strong> <strong>METHOD</strong> (Q-<strong>HAM</strong>)<br />
Consider the following differential equation:<br />
[ ] (1)<br />
where N is a nonlinear operator, denotes independent variables, is a known<br />
function and is an unknown function.<br />
Let us construct the so-called zero-order deformation equation:<br />
[ ] [ ] (2)<br />
Where , [ ] denotes the so-called embedded parameter , is an auxiliary linear<br />
operator with the property [ ] , is an auxiliary parameter,<br />
denotes a non-zero auxiliary function.<br />
It is obvious that when<br />
respectively. Thus as increases from 0 to<br />
guess to the solution .<br />
(<br />
equation (2) becomes:<br />
) (3)<br />
, the solution varies from the initial<br />
Having the freedom to choose , we can assume that all of them can be<br />
properly chosen so that the solution of equation (2) exists for [ ].<br />
Expanding in Taylor series, one has:<br />
Where:<br />
∑<br />
(4)<br />
(5)
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Assume that are so properly chosen such that the series (4) converges at<br />
and:<br />
(<br />
) ∑ (<br />
Defining the vector { }<br />
differentiating equation (2) times with respect to and then setting and finally<br />
dividing them by we have the so-called order deformation equation:<br />
[ ]<br />
where :<br />
and:<br />
{<br />
[ ]<br />
It should be emphasized that for is governed by the linear equation (7) with<br />
linear boundary conditions that come from the original problem. Due to the existence of the<br />
factor (<br />
)<br />
)<br />
, more chances for convergence may occur or even much faster convergence can<br />
be obtained better than the standard <strong>HAM</strong>. It should be noted that the case of<br />
, standard <strong>HAM</strong> can be reached.<br />
3 APPLICATIONS<br />
3.1 The nonlinear homogeneous gas dynamics equation<br />
Let<br />
with initial condition<br />
The exact solution of this problem is known to be<br />
This problem was solved by <strong>HAM</strong> in (Hany Nasr Hassan 2010) .<br />
For q- <strong>HAM</strong> solution we choose the linear operator :<br />
[ ]<br />
with the property :<br />
[ ]<br />
(6)<br />
(7)<br />
(8)<br />
(9)<br />
(10)<br />
(11)<br />
(12)<br />
(13)<br />
53
54<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
where is constant. Using initial approximation ,<br />
we define a nonlinear operator as<br />
[ ]<br />
We construct the zero order deformation equation:<br />
[ ] [ ]<br />
we can take , and the order deformation equation is :<br />
[ ]<br />
with the initial conditions for<br />
Where:<br />
And:<br />
∑<br />
{<br />
∑<br />
[ ]<br />
Now the solution of equation (10) for becomes<br />
∫<br />
where the constant of integration is determined by the initial conditions (15).<br />
We can obtain components of the solution using q- <strong>HAM</strong> as follows:<br />
can be calculated similarly. Then the series solution expression by<br />
q- <strong>HAM</strong> can be written in the form:<br />
∑ (<br />
)<br />
(14)<br />
(15)<br />
(16)<br />
,
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Equation (16) is an approximate solution to the problem (10) in terms of the convergence<br />
parameters . It can be noted that the fraction factor ( )<br />
convergence chances than that of <strong>HAM</strong>.<br />
55<br />
highly increase the<br />
To find the valid region of , the -curves given by the 10 th order q-<strong>HAM</strong> approximation at<br />
different values of are drawn in figures ). These figures show the<br />
interval of at which the value of is constant at certain values of .<br />
We choose the horizontal line parallel to as a valid region which provides us<br />
with a simple way to adjust and control the convergence region of the series solution(16).<br />
From these figures , the valid intersection region of for the values of in the<br />
curves becomes larger as increase as in the following Table(1):<br />
10<br />
20<br />
30<br />
40<br />
50<br />
100<br />
region<br />
Table (1): the increase of the convergence interval length with the increase of n<br />
Figures show the comparison among using different values of with<br />
the exact solution (12). Figures show the comparison between of <strong>HAM</strong><br />
and of q-<strong>HAM</strong> using different values of with the exact solution (12), which indicates<br />
that the speed of convergence for q-<strong>HAM</strong> with is faster in comparison with .<br />
The Absolute errors of the 10 th order solutions q-<strong>HAM</strong> approximate at using different<br />
values of compared with 10 th order solutions <strong>HAM</strong> approximate at are<br />
calculated by the formula<br />
| | (17)<br />
Figures show that the series solution obtained by <strong>HAM</strong> is more accurate at<br />
but at larger the series solutions obtained by q-<strong>HAM</strong> at converge faster than<br />
(<strong>HAM</strong>).
56<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
2.5 2.0 1.5 1.0 0.5 0.5 h<br />
Figure : - curve for the <strong>HAM</strong> (q-<strong>HAM</strong>; approximation solution of<br />
problem (10) at different values of .<br />
3 2 1<br />
U 10 x, t , n<br />
40<br />
30<br />
20<br />
10<br />
10<br />
20<br />
U 10 x, t , n<br />
30<br />
20<br />
10<br />
U 10 0,1,1<br />
U 10 0.5 ,2,1<br />
U 10 0.7 ,3,1<br />
U 10 0.1 ,3.7 ,1<br />
Figure : - curve for the ( q-<strong>HAM</strong>; approximation solution of<br />
problem (10) at different values of .<br />
h<br />
U 10 0,1,1.5<br />
U 10 0.5 ,2,1.5<br />
U 10 0.7 ,3,1.5<br />
U 10 0.1 ,3.7 ,1.5
12 10 8 6 4 2<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Figure : - curve for the (q-<strong>HAM</strong>; approximation solution of problem<br />
(10) at different values of .<br />
25 20 15 10 5<br />
U 10 x, t , n<br />
Figure : - curve for the (q-<strong>HAM</strong>; approximation solution of<br />
problem (10) at different values of .<br />
30<br />
20<br />
10<br />
U 10 x, t , n<br />
40<br />
30<br />
20<br />
10<br />
10<br />
20<br />
30<br />
h<br />
h<br />
U 10 0,1,5<br />
U 10 0.5 ,2,5<br />
U 10 0.7 ,3,5<br />
U 10 0.1 ,3.7 ,5<br />
U 10 0,1,10<br />
U 10 0.5 ,2,10<br />
U 10 0.7 ,3,10<br />
U 10 0.1 ,3.7 ,10<br />
57
58<br />
solutions<br />
60000<br />
50000<br />
40000<br />
30000<br />
20000<br />
10000<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
2 4 6 8 10 12 14<br />
Figure : Comparison between of <strong>HAM</strong> (q-<strong>HAM</strong>; and exact solution of<br />
(10) at with ,<br />
solutions<br />
80000<br />
60000<br />
40000<br />
20000<br />
2 4 6 8 10 12 14<br />
Figure : Comparison between of (q-<strong>HAM</strong> and exact solution of (10) at<br />
with ,<br />
t<br />
t<br />
Exact<br />
U 10 n 1<br />
U 7 n 1<br />
U 5 n 1<br />
Exact<br />
U 10 n 5<br />
U 7 n 5<br />
U 5 n 5
250000<br />
200000<br />
150000<br />
100000<br />
50000<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
2 4 6 8 10 12 14<br />
Figure : Comparison between of <strong>HAM</strong> (q-<strong>HAM</strong>;( )) and (q-<strong>HAM</strong><br />
with exact solution of (10) at with<br />
5 10 6<br />
4 10 6<br />
3 10 6<br />
2 10 6<br />
1 10 6<br />
solutions<br />
solutions<br />
5 10 15 20<br />
Figure : Comparison between of <strong>HAM</strong> (q-<strong>HAM</strong> and q-<strong>HAM</strong>(<br />
with exact solution of (10) at with (<br />
,<br />
t<br />
t<br />
Exact<br />
U10 n 1<br />
U10 n 1.5<br />
U10 n 2<br />
U10 n 3<br />
U10 n 4<br />
U10 n 5<br />
Exact<br />
U10 n 1<br />
U10 n 10<br />
U10 n 20<br />
U10 n 30<br />
U10 n 40<br />
U10 n 50<br />
U10 n 100<br />
59
60<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
Figure : The Absolute error of of q-<strong>HAM</strong> ( for problem (10) at<br />
and using<br />
Figure : The Absolute error of of q-<strong>HAM</strong> ( for problem (10) at<br />
and using<br />
3.2 The Riccati equation<br />
Let<br />
Absolute Error<br />
0.0012<br />
0.0010<br />
0.0008<br />
0.0006<br />
0.0004<br />
0.0002<br />
Absolute Error<br />
30 000<br />
25 000<br />
20 000<br />
15 000<br />
10 000<br />
5000<br />
The exact solution is known to be<br />
1 2 3 4<br />
8 10 12<br />
t<br />
t<br />
A.E n 1<br />
A.E n 40<br />
A.E n 1<br />
A.E n 40<br />
(18)<br />
(19)
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
For q- <strong>HAM</strong> solution we choose the linear operator :<br />
[ ]<br />
With the property :<br />
[ ]<br />
Where is constant. Using initial approximation<br />
We define a nonlinear operator as :<br />
[ ]<br />
We construct the zero order deformation equation<br />
[ ] [ ]<br />
Where . We can take H , and the order deformation equation is :<br />
[ ]<br />
(21)<br />
With the initial conditions for :<br />
Where:<br />
and<br />
{<br />
∑<br />
[ ]<br />
Now the solution of equation (18) for becomes:<br />
∫<br />
Where the constant of integration is determined by the initial conditions (22).<br />
We now obtain components of the solution using q- <strong>HAM</strong> as follows:<br />
,<br />
can be calculated similarly.Then the series solution expression by q-<br />
<strong>HAM</strong> can be written in the form:<br />
(20)<br />
(22)<br />
61
62<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
∑ (<br />
Equation (23) is a family of approximation solutions to the problem (18) in terms of the<br />
convergence parameters .<br />
)<br />
To find the valid region of , the -curves given by the 15 th order q-<strong>HAM</strong> approximation at<br />
different values of are drawn in figures . From these figures , the valid<br />
intersection region of for the values of in the curves becomes larger as increase<br />
as in the following Table(2).<br />
10<br />
20<br />
30<br />
40<br />
50<br />
100<br />
region<br />
Table (2): the increase of the convergence interval length with the increase of n<br />
U 15 x, n<br />
2.0 1.5 1.0 0.5 0.5 h<br />
10<br />
5<br />
5<br />
U 15 0.1 ,1<br />
U 15 0.4 ,1<br />
U 15 0.6 ,1<br />
U 15 0.8 ,1<br />
Figure : - curves for the <strong>HAM</strong> (q-<strong>HAM</strong>; approximation solution of<br />
problem (18) at different values of .<br />
(23)
200 150 100 50<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Figure : - curves for the (q-<strong>HAM</strong>; approximation solution of<br />
problem (18) at different values of .<br />
Figures show the comparison between of <strong>HAM</strong> and of q-<strong>HAM</strong> using<br />
different values of with the exact solution (19), which indicates that the speed of<br />
convergence for q-<strong>HAM</strong> with is faster in comparison with .<br />
solutions<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
U 15 x, n<br />
15<br />
10<br />
0.2 0.2 0.4 0.6 0.8 1.0<br />
Figure : Comparison between of <strong>HAM</strong>(q-<strong>HAM</strong> and q-<strong>HAM</strong>( with<br />
exact solution of (18) with( .<br />
5<br />
x<br />
h<br />
U 15 0.1 ,100<br />
U 15 0.4 ,100<br />
U 15 0.6 ,100<br />
U 15 0.8 ,100<br />
Exact<br />
U 15 n 2<br />
U 15 n 1<br />
63
64<br />
solutions<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
0.2 0.2 0.4 0.6 0.8 1.0<br />
Figure : Comparison between of <strong>HAM</strong>(q-<strong>HAM</strong> and q-<strong>HAM</strong>(<br />
with exact solution of (18) with( .<br />
The absolute errors of the 15 th order solutions q-<strong>HAM</strong> approximate using different values of<br />
compared with 15 th order solutions <strong>HAM</strong> approximate are shown in figures (15,16).<br />
These figures show that the series solution obtained by <strong>HAM</strong> is more accurate at<br />
but at larger the series solution obtained by q-<strong>HAM</strong> at converge faster than<br />
(<strong>HAM</strong>).<br />
Absolute Error<br />
0.00005<br />
0.00004<br />
0.00003<br />
0.00002<br />
0.00001<br />
0.1 0.2 0.3 0.4 0.5<br />
Figure : The Absolute error of of q-<strong>HAM</strong> ( for problem (18) at<br />
and using<br />
x<br />
x<br />
Exact<br />
U 15 n 100<br />
U 15 n 1<br />
A.E n 1<br />
A.E n 100
Absolute Error<br />
0.25<br />
0.20<br />
0.15<br />
0.10<br />
0.05<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
0.60 0.65 0.70 0.75 0.80 0.85<br />
Figure : The Absolute error of of q-<strong>HAM</strong> ( for problem (18) at<br />
and using<br />
It should be noted that the absolute error is highly decreased when modifying the solution by<br />
taking more terms into consideration. Figures (17,18) illustrate this fact.<br />
Absolute Error<br />
0.0008<br />
0.0006<br />
0.0004<br />
0.0002<br />
0.1 0.2 0.3 0.4 0.5<br />
Figure : The Absolute error of ( and of q-<strong>HAM</strong> at<br />
using respectively.<br />
x<br />
x<br />
A.E n 1<br />
A.E n 100<br />
A.E U 10 n 1<br />
A.E U 10 n 100<br />
A.E U 30 n 100<br />
65
66<br />
Absolute Error<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
Figure : The Absolute error of ( and of q-<strong>HAM</strong><br />
at using respectively.<br />
3.3 The logistic growth model<br />
where r and K are positive constants. Here represents the population of the species<br />
at time , and K is the carrying capacity of the environment.<br />
Let :<br />
1.2<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
Then (1) becomes<br />
If then<br />
therefore, the analytical solution of (25) is easily obtained<br />
(such as Bernoulli method and separation of variables)<br />
(26)<br />
For q- <strong>HAM</strong> solution we choose the linear operator :<br />
[ ]<br />
With the property :<br />
[ ]<br />
0.6 0.7 0.8 0.9<br />
Where is constant.Using initial approximation :<br />
x<br />
A.E U 10 n 1<br />
A.E U 10 n 100<br />
A.E U 30 n 100<br />
(24)<br />
(25)<br />
(27)<br />
(28)
We define a nonlinear operator as :<br />
[ ]<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
We construct the zeroth-order deformation equation:<br />
[ ] [ ]<br />
We can take , and the mth-order deformation equation is :<br />
[ ]<br />
With the initial conditions for :<br />
Where:<br />
And:<br />
{<br />
[ ]<br />
Now the solution of equation (29) for becomes:<br />
∫<br />
∑<br />
2here the constant of integration is determined by the initial conditions (30).<br />
For numerical results we take , therefore (26) becomes:<br />
We now obtain components of the solution using q- <strong>HAM</strong> as follows:<br />
can be calculated similarly.Then the series solution expression by q-<br />
<strong>HAM</strong> can be written in the form:<br />
∑ (<br />
Equation (33) is a family of approximation solutions to the problem (24) in terms of the<br />
convergence parameters .<br />
)<br />
(29)<br />
(30)<br />
(31)<br />
(32)<br />
(33)<br />
67
68<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
To find the valid region of ,the -curves given by the 10 th order q-<strong>HAM</strong> approximation at<br />
different values of are drawn in figures (19 ,20 and 21).This figures show the<br />
interval of where the value of is constant at certain . We choose the<br />
horizontal line parallel to as a valid region which provides us with a simple way<br />
to adjust and control the convergence region.<br />
5 10 14<br />
4 10 14<br />
3 10 14<br />
2 10 14<br />
1 10 14<br />
U 10 , n<br />
0.06 0.04 0.02 0.02 0.04 0.06<br />
Figure : - curves for the <strong>HAM</strong> (q-<strong>HAM</strong>; approximation solution of<br />
problem (1) at different values of .<br />
U 10 , n<br />
2000<br />
1500<br />
1000<br />
500<br />
0.06 0.04 0.02 0.02<br />
Figure : - curves for the q-<strong>HAM</strong>( approximation solution of<br />
problem (1) at different values of .<br />
h<br />
h<br />
U 10 0.1 ,1<br />
U 10 0.6 ,1<br />
U 10 1,1<br />
U 10 2,1<br />
U 10 0.1 ,10<br />
U 10 0.6 ,10<br />
U 10 1,10<br />
U 10 2,10
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
0.30 0.25 0.20 0.15 0.10 0.05 0.00 0.05<br />
Figure : - curves for the q-<strong>HAM</strong>( approximation solution of<br />
problem (1) at different values of .<br />
Figures (22,23 and 24) show the comparison between of <strong>HAM</strong> and of q-<strong>HAM</strong> using<br />
different values of with the solution (32)<br />
solutions<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
U 10 , n<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
Figure(22): Comparison between of <strong>HAM</strong>(q-<strong>HAM</strong> and q-<strong>HAM</strong>( and<br />
for logistic growth model (1) with( respectively.<br />
h<br />
U 10 0.1 ,100<br />
U 10 0.6 ,100<br />
U 10 1,100<br />
U 10 2,100<br />
U<br />
U 10 n 2<br />
U 10 n 1<br />
69
70<br />
solutions<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
Figure(23): Comparison between of <strong>HAM</strong>(q-<strong>HAM</strong> and q-<strong>HAM</strong>( and<br />
for logistic growth model (1) with( respectively.<br />
solutions<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
Figure(24): Comparison between of <strong>HAM</strong>(q-<strong>HAM</strong> and q-<strong>HAM</strong>( and<br />
for logistic growth model (1) with( respectively.<br />
The Absolute errors of the 10 th order solutions q-<strong>HAM</strong> approximate using different values of<br />
compared with 10 th order solutions <strong>HAM</strong> approximate are calculated by the formula<br />
| | .<br />
U 10 n 10<br />
U 10 n 1<br />
Figures show that the series solution obtained by <strong>HAM</strong> is divergent at<br />
but, the series solutions obtained by q-<strong>HAM</strong> becomes more accurate<br />
as increases.<br />
U<br />
U<br />
U 10 n 100<br />
U 10 n 1
Absolute Error<br />
140<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
Figure(25): The Absolute error of of <strong>HAM</strong> (q-<strong>HAM</strong> ( and q-<strong>HAM</strong> ( for<br />
problem (1) using with<br />
Absolute Error<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
A.E n 2<br />
A.E n 1<br />
A.E n 10<br />
A.E n 1<br />
Figure(26): The Absolute error of of <strong>HAM</strong> (q-<strong>HAM</strong> ( and q-<strong>HAM</strong> ( for<br />
problem (1) using with .<br />
71
72<br />
Absolute Error<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
Figure(27): The Absolute error of of <strong>HAM</strong> (q-<strong>HAM</strong> ( and q-<strong>HAM</strong> ( for<br />
problem (1) using with<br />
4 CONCLUSION<br />
A more general method (q-<strong>HAM</strong>) of the homotopy analysis method (<strong>HAM</strong>) was proposed.<br />
This method provides an approximate solution by taking different values of . The results<br />
show that the convergence region of series solutions obtained by q-<strong>HAM</strong> is increasing as q is<br />
decreased. The q-<strong>HAM</strong> improves the performance of <strong>HAM</strong> and it was shown from the<br />
illustrative examples that the convergence of q-<strong>HAM</strong> is faster than the convergence of <strong>HAM</strong>.<br />
REFERENCES<br />
0.5 1.0 1.5 2.0 2.5 3.0<br />
A.E n 100<br />
A.E n 1<br />
Adomain G and Adomain GE(1984).A global method for solution of complex systems.<br />
Mathematical Modelling ;5;521-68.<br />
Adomain G(1991).A review of the decomposition method and some recent results for<br />
nonlinear equations. Computers and Mathematics with Applications ;21:101-27.<br />
Awrejcewicz J Andrianov IV Manevitch LI(1998). Asymptotic approaches in nonlinear<br />
dynamics. Berlin:Springer-Verlag.<br />
AK Alomari Mohd Salmi Md Noorani Roslinda Mohd Nazar(2008). Approximate analytical<br />
solutions of the Klein-Gordon equation by means of the homotopy analysis method, Journal<br />
of Quality Measurement and Analysis 4(1),45-57.<br />
A Sami Bataineh I Hashim S Abbasbandy MSM Noorani(2011). Direct solution of singular<br />
higher-order BVPs by the homotopy analysis method and its modification, World Applied<br />
Sciences Journal 12(6):884-889.<br />
Bataineh AS Noorani MSM Hashim I(2007). Solutions of time-dependent Emden-Fowler<br />
type equations by homotopy analysis method. Physics Letters A ;371:72-82.
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Bataineh AS Noorani MSM Hashim I(2009). On a new reliable modification of homotopy<br />
analysis method. Communications in Nonlinear Science and Numerical Simulation ,14:409-<br />
423.<br />
Chin Fung Yuen Lem Kong Hong Chong Fook Seng(2010). Solving nonlinear algebraic<br />
equation by homotopy analysis method ,proceedings of the Gth IMT-GT conference on<br />
mathematics ,statistics and its applications(ICMSA) University Tunku Abdul Rahman,Kuala<br />
Lumpur,Malaysia.<br />
Hany Nasr Hassan Mohammed(2010),The application of homotopy analysis method in<br />
solving initial and boundary value problems., Ph.D. thesis ,Faculty of engineering, Cairo<br />
University.<br />
Hany N Hassan and Magdy A El-Tawil(2011). A new technique of using homotopy analysis<br />
method for solving high-order nonlinear differential equations. , Math. Meth. Appl. Sci, 34<br />
:728–742.<br />
Hany N Hassan and Magdy A El-Tawil(2011). An efficient analytic approach for solving<br />
two-point nonlinear boundary value problems by homotopy analysis method , Math. Meth.<br />
Appl. Sci. 34: 977–989.<br />
Hany N Hassan and Magdy A El-Tawil(2011). SOLVING CUBIC AND COUPLED<br />
NONLINEAR SCHRÖ DINGER EQUATIONS USING <strong>THE</strong> <strong>HOMOTOPY</strong> <strong>ANALYSIS</strong><br />
<strong>METHOD</strong>, Int. J. of Appl. Math and Mech. 7 (8): 41-64<br />
.<br />
Hany N Hassan and Magdy A El-Tawil(2011). MODIFIED <strong>HOMOTOPY</strong> <strong>ANALYSIS</strong><br />
<strong>METHOD</strong> FOR SECOND ORDER NONLINEAR INITIAL VALUE PROBLEMS, Int. J. of<br />
Appl. Math and Mech. 7 (18): 82-108.<br />
Hinch EJ(1991).Perturbation methods .Cambridge Texts in Applied mathematics. Cambridge<br />
university Press.<br />
H Hossein Zadeh H Jafari and S M Karimi(2010). Homotopy analysis method for solving<br />
integral and integro_differential equations, International Journal of Research and Reviews in<br />
Applied Sciences2(2).February .<br />
Karmishin AV Zhukov AT Kolosov VG(1990). Methods of dynamics calculation and testing<br />
for thin-walled structures ,Moscow:Mashinostroyenie [in Russian].<br />
Lagerstrom PA(1988). Matched asymptotic expansions,Ideas and techniques of applied<br />
mathematical sciences,vol.76.New York:Springer-Verlag.<br />
Liao SJ(1992).The proposed homotopy analysis techniques for the solution of nonlinear<br />
problems. Ph.D. thesis, Shanghai Jiao Tong University.<br />
Liao SJ(1995). An approximate solution technique not depending on small parameters: A<br />
special example .International Journal of non-linear mechanics ;30(2):371-380.<br />
73
74<br />
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
M. A. El-Tawil and S. N. Huseen<br />
Liao SJ(1996).What’s the common ground of all numerical and analytical tichniques for<br />
nonlinear problems.Communications in nonlinear science & numerical simulation; (4):26-30.<br />
Lioa SJ(1997). Homotopy analysis method :A new analytical technique for nonlinear<br />
problems. Communications in nonlinear science & numerical simulation; 2(2):95-100.<br />
Lioa SJ(2002). A direct boundary element approach for unsteady non-linear heat transfer<br />
problems. Engineering analysis with boundary element method ;26:55-59.<br />
Lioa SJ(2003). Beyond perturbation : Introduction to the homotopy analysis method .CRC<br />
Press,Chapman & Hall.<br />
Loia SJ(2004). On the homotopy analysis method: for nonlinear problems .Applied<br />
mathematics and computation ; 147:499-513.<br />
Liao SJ(2009). Notes on the homotopy analysis method : Some definitions and theorems .<br />
Communications in nonlinear science & numerical simulation ,14:983-997.<br />
Lyapunov AM ( 1992). General problem on stability of motion .Taylor & Francis ,London<br />
[English translation].<br />
Murdock JA(1991).Perturbation :theory and methods.New York:John Wiley&Sons.<br />
Nayfeh AH(1981). Introduction to perturbation techniques ,New York; John Wiley &Sons.<br />
Nayfeh AH(1985). Problems in perturbation ,New York; John Wiley &Sons.<br />
Nayfeh AH(2000). Perturbation methods. New York: John Wiley&Sons.<br />
O Abdulaziz I Hashim and A Saif(2008). Series Solutions of Time-Fractional PDEs by<br />
Homotopy Analysis Method, Hindawi Publishing Corporation Differential Equations and<br />
Nonlinear Mechanics Volume 2008, Article ID 686512, 16 pages, doi:10.1155/686512.<br />
Rach R(1984). On the Adomain method and comparisons with Picards method . Journal of<br />
Mathematical Analysis and Applications ;10:139-59.<br />
Seyyed Ali Kazemipour and Ahmad Neyrameh(2010).Analytical study on Goursat<br />
problems,world applied sciences Journal;10(8) 867-870.<br />
Song H and Tao L(2007). Homotopy analysis of 1 D unsteady, nonlinear groundwater flow<br />
through porous media. Journal of Coastal Research ;50:292-5.<br />
Sukran Cavdar(2010). Homotopy analysis method for one group Neutron diffusion equation,<br />
5 th International Ege Energy Symposium and Exhibition (IEESE-5) 27-30 June,Pamukkale<br />
University , Denizli, Turkey.<br />
Tao L Song H Chakrabarti S(2007). Nonlinear progressive waves in water of finite depth-an<br />
analytic approximation,Coastal Engineering ;54:825-34.
Int. J. of Appl. Math and Mech. 8 (15): 51-75, 2012.<br />
The q-Homotopy Analysis Method<br />
Van Gorder RA. and Vajravelu K(2009). On the selection of auxiliary functions, operators<br />
,and convergence control parameters in the application of the homotopy analysis method to<br />
nonlinear differential equations:A general approach. Communications in nonlinear science &<br />
numerical simulation; 14:4078-4089<br />
Wang H Zou L Zhang H(2007).Applying homotopy analysis method for solving differential<br />
difference equation. Physics Letters A; 369:77-84.<br />
75