Appl. Math. Inf. Sci. 8, No.

5, 2171-2176 (2014) 2171

Applied Mathematics & Information Sciences
An International Journal

http://dx.doi.org/10.12785/amis/080510

On Finding Exact and Approximate Solutions to Some

PDEs Using the Reduced Differential Transform Method
Mahmoud Rawashdeh∗ and Nazek A. Obeidat
Department of Mathematics and Statistics, Jordan University of Science and Technology, Irbid 22110, Jordan

Received: 13 Aug. 2013, Revised: 10 Nov. 2013, Accepted: 11 Nov. 2013

Published online: 1 Sep. 2014

Abstract: In this research article, we give analytic approximate solution to the Sharma Tasso Olver (STO) equation and exact solutions
to both the Schrodinger equation and the Telegraph equation. Also, the approximate analytical and exact solutions we present in this
paper are calculated in the form of power series with easily computable components. The obtained results are in a good agreement with
the exact solutions. We present an algorithm called the Reduced Differential Transform Method (RDTM) to find approximate solution
and we compare the results with the exact solutions. This method reduces significantly the numerical computations compare with the
existing methods such as the perturbation technique, differential transform method (DTM) and the Adomian decomposition method
(ADM).

Keywords: Reduced Differential Transform Method (RDTM), Sharma Tasso Olver (STO) equation, Schrodinger equation, Telegraph
equation, Approximate solutions, Analytical solutions.

1 Introduction solutions with high accuracy. In addition, M. Rawashdeh

[14] used the RDTM to find exact and approximate
solution for Gardner equation, Variant Nonlinear Water
There are many wave equations which are quite useful in Wave equation (VNWW), and the Fifth-Order
physics and engineering. These equations are represented Korteweg-de Vries (FKdV) equation. Finally, Abazari. R,
usually by linear and nonlinear PDEs and solving such Soltanalizadeh [16] used the RDTM to find approximate
equations is very important and sometime it is difficult to solutions for the Kawahara Equations.
handle the nonlinear part of these PDEs. Many authors
applied numerical methods to find solutions of these In this paper, we apply the RDTM to the Sharma Tasso
equations and to name few of these methods: The Olver (STO) equation which is a good example to show
Differential Transform Method (DTM) [7, 12], the fission and fusion of the soliton solutions. The SOT was
Adomian Decomposition Method (ADM) [5, 6], the studied by many authors using different methods such as
Variational Iteration Method (VIM) [8, 11] and the Hirota’s direct method [17] and extended tanh method
sine–cosine method [9]. The RDTM was first introduced [10]. Also, A. M. Wazwaz [10] found solitons and kinks
by Y. Keskin in his Ph.D. [4]. This method based on the solutions to the Sharma Tasso Olver equation.
use of the traditional DTM techniques. Usually, a few The standard form of the Sharma Tasso Olver equation
numbers of iteration needed of the series solution for [10, 17] is given by
numerical purposes with high accuracy.
3
ut + α u3 + α (u2 )xx + α uxxx = 0,


The RDTM has been used by many authors to obtain x
(1)
2
analytical approximate and in some cases exact solutions
to nonlinear wave equations. Keskin and Oturanc [1, 2, 3] where α is a constant.
used the RDTM to solve linear and nonlinear wave
equations and they showed the effectiveness, and the In this paper, we present analytic approximate solution to
accuracy of the proposed method. Moreover, they showed Eq.(1) and exact solutions for both Eq.(4) and Eq.(6).
that it takes only few iterations to get an approximate
∗ Corresponding author e-mail: msalrawashdeh@just.edu.jo
c 2014 NSP
Natural Sciences Publishing Cor.
2172 M. Rawashdeh, N. Obeidat: On Finding Exact and Approximate Solutions to Some...

First, the Sharma Tasso Olver equation of the form:

In this work, the lowercase u (x,t) represent the original
3 function while the uppercase Uk (x) stand for the
ut + α u x + α (u2 )xx + α uxxx = 0,
3


(2)
2 transformed function. Note that from the above
discussion, one can realize that the RDTM is derived
subject to the initial condition
from the power series expansion.
r r ! Some basic operations of the reduced differential
1 1
u(x, 0) = tanh x . (3) transformation obtained from equations (8) and (9) are
α α given in the table below:

Second, the Schrodinger equation of the form: Table 1. Basic operations of the RDTM [1, 2, 3, 4]
ut = iuxx , (4) Functional Form Transformed  form 
u (x,t) 1 ∂ k u(x,t)
Uk (x) = k!
∂ tk t=0
w (x,t)=α u(x,t) ± β v(x,t) Wk (x) = α Uk (x) ± β Vk (x), α and β are constants.
subject to the initial condition w (x,t) = u(x,t)v(x,t) Wk (x) = ∑ki=0 Ui (x)Vk−i (x)
f (x,t)=u(x,t)v(x,t)w(x,t) Fk (x) = ∑ki=0 ∑ij=0 U j (x)Vi− j (x)Wk−i (x)
u(x, 0) = sinh(x), (5) n
w (x,t) = ∂ n u(x,t)
∂t
(k+n)!
Wk (x) = K! Uk+n (x)
n n
w (x,t) = ∂ n u(x,t) Wk (x) = ∂ n Uk (x)
∂x ∂x
Third, the homogeneous Telegraph equation of the form: w (x,t) = xm t n u(x,t) Wk (x) = xm Uk−n (x)

1, n=k
w (x,t) = xm t n Wk (x)=xm δ (k − n), where δ (k-n)=
0, n6=k
uxx = utt − 2ut − u, (6) n+m
w (x,t)= ∂ n m u(x,t)
n
Wk (x)= ∂ n

(k+m)!

∂x ∂t ∂x k! Uk+m (x)

subject to the initial condition

Remark. It is worth mentioning here that table 1 was
u(x, 0) = cosh(x) − 1; ut (x, 0) = 1. (7)
derived by Y. Keskin in his Ph.D [4]. The proofs of theses
theorems are also available in [4].
The rest of this paper is organized as follows: In Section
2, the reduced differential transform method is Now, we illustrate the RDTM by using Eq.(1) in standard
introduced. Section 3 is devoted to apply the RDTM to form
three test problems to show the effectiveness of the
RDTM. In section 4, we present a table to show the L (u(x,t)) + R (u(x,t)) + N (u(x,t)) = 0 (11)
comparison between the approximate and exact solutions
using the RDTM. Section 5 discussion and conclusion of with initial conditions
this paper.
u(x, 0) = f (x); ut (x, 0) = g(x), (12)
where L = ∂∂t is a linear operator,
2 Analysis of the Method
N (u(x,t)) = α u3 x + 32 α (u2 )xx is the nonlinear term and

In this section, we will give the methodology of the R (u(x,t)) is the remaining linear term.
RDTM. So let’s start with a function of two variables
u (x, t) which is analytic and k−times continuously Using the RDTM formulas in Table 1, we can derive the
differentiable with respect to time t and space x in the following recursive relation:
domain of our interest. Assume we can represent this
function as a product of two single-variable functions (k + 1)Uk (x) = R (Uk (x)) − N (Uk (x)) +Uk (x) (13)
u (x, t) = f (x).g(t). From the definitions of the DTM, the
function can be represented as follows: where, R (Uk (x)), Uk (x) and N (Uk (x)) are the
transformations of R (u(x,t)), u(x,t) and N (u(x,t))
∞ ∞ ∞
! !
respectively.
u (x,t) = ∑ F(i)x i
∑ G( j)t j
= ∑ Uk (x)t k (8) Now from equation (12), we can write the initial
i=0 j=0 k=0
condition as:
where Uk (x) is the transformed function of u (x,t) which U0 (x) = f (x); U1 (x) = g(x) (14)
can be defined as:
1 ∂k
 
To find all other iterations, we first substitute equation (14)
Uk (x) = u(x,t) . (9)
k! ∂ t k t=0
into equation (13) and then we find the values of Uk (x).
Finally, we apply the inverse transformation to all values
From equations (8) and (9) one can deduce {Uk (x)}nk=0 to obtain the approximate solution:
∞ ∞ n
1 ∂k
∑ Uk (x)t k .
 
⌢
u (x,t) = ∑ k! ∂ t k u(x,t) t k = ∑ Uk (x)t k . (10) u (x,t) = (15)
k=0 t=0 k=0 k=0

c 2014 NSP
Natural Sciences Publishing Cor.
Appl. Math. Inf. Sci. 8, No. 5, 2171-2176 (2014) / www.naturalspublishing.com/Journals.asp 2173

We continue in this manner and after a few iterations, the

where n is the number of iterations we need to find the differential inverse transform of {Uk (x)}∞
k=0 will provide
intended approximate solution. us with the following approximate solution:
Hence, the exact solution of our problem is given by ∞
∑ Uk (x)t k
⌢ ⌢
u(x,t) = lim u (x,t). u (x,t) =
n→∞ k=0
= U0 (x) +U1 (x)t +U2 (x)t 2 +U3 (x)t 3 + ....
1 x 1 x sinh(x)
3 Numerical Examples = tanh − sech2 t− t2
2 2 4 2 4(1 + cosh(x))2
In this section, we apply the RDTM to three numerical 1 x
− (cosh(x) − 2)sech4 t3 + . . . .
examples and then compare our approximate solutions to 48 2
the exact solutions to show the efficiency of the RDTM.
Hence, the approximate solution converges rapidly to the
exact solution and the exact solution of the problem is
⌢
3.1 Sharma Tasso Olver (STO) equation given by u(x,t) = lim u n (x,t).
n→0

First, consider the Sharma Tasso Olver (STO) equation: From figure 1 below one can observe that the values of
the approximate solution at different grid points obtained
3 by RDTM are very close to the values of the exact
ut + α u3 + α (u2 )xx + α uxxx = 0,


x
(16) solution with high accuracy with only five iterations and
2
the accuracy increases as the order of approximation
where α is a constant. increases.

In the case when α = 4, the STO becomes

ut + 4 u3 x + 6(u2 )xx + 4uxxx = 0,

(17)
0.5 0.5
0.01 0.01 4. ´ 10-16
0.0 0.0
subject to the initial conditions -0.5
-5 t
0.008
0.006 -0.5
-5
0.008 2. ´ 10-16
0.006
t 0 0.006
t
0.01
0.008

-3 0.004 -3 0.004 -5
0.004
-3
0 0.002 0 0.002 0 0.002
x x
−1
x
1 x x 3 0 3 0 3 0

sech2
5 5 5
u(x, 0) = tanh ; ut (x, 0) = , (18)
2 2 4 2
Fig. 1: The approximate, exact solutions and absolute error, respectively for example 3.1 when −5 < x < 5
and 0 < t < 0.01
where the exact solution is
 
1 x−t
u(x,t) = tanh . (19) Also figure 2 below shows the exact solution,
2 2
approximate solution of u(x,t) for the values of
Applying the RDTM to (18) and (17), we obtain the x = −5, −3, 3, 5 and t = 0.02, 0.04, 0.06, 08, 0.1.
recursive relation
  
−1  ∂ 3 Uk (x) ∂  k i
∂ x ∑ ∑ i− j
Uk+1 (x) = 4 +4 U (x)U j (x)Uk−i (x)
k+1 ∂ x3 i=0 j=0 approximate exact

!!
−6 ∂2 k 0.4 0.4
+
k + 1 ∂ x2 ∑ i k−i (x) .
U (x)U (20)
i=0 0.2 0.2

where the Uk (x), is the transform function of the -4 -2 2 4

x
-4 -2 2 4
x

t−dimensional spectrum. Note that -0.2 -0.2

−1
-0.4 -0.4
1 x x
U0 (x) = tanh ; U1 (x) = sech2 . (21)
2 2 4 2
Fig. 2: The approximate and exact solutions for example 3.1 when −5 < x < 5 and 0 < t < 0.1
Now, substitute Eq. (21) into Eq. (20) to obtain the
following:
sinh(x)
U2 (x) = 4(1+cosh(x))2 3.2 Schrodinger equation
1
(cosh(x) − 2)sech4 2x


U3 (x) = − 48
Consider the Schrodinger equation of the form:
(cosh(x)−5)tanh( x
2 )
U4 (x) = − 48(1+cosh(x))2
. ut = i uxx , (22)

c 2014 NSP
Natural Sciences Publishing Cor.
2174 M. Rawashdeh, N. Obeidat: On Finding Exact and Approximate Solutions to Some...

subject to the initial condition where the Uk (x), is the transform function of the
t−dimensional spectrum. Note that
u(x, 0) = sinh(x), (23)
U0 (x) = cosh(x) − 1; U1 (x) = 1. (31)
where the exact solution is Now, substitute Eq. (30) into Eq. (31) to obtain the
it following:
u(x,t) = e sinh(x). (24)
−1 1
Applying the RDTM to (22) and (23), we obtain the U2 (x) = , U3 (x) = , ... (32)
recursive relation 2 6

i ∂
 2  So after a few iterations, the differential inverse transform
Uk+1 (x) = (Uk (x)) , (25) of {Uk (x)}∞ k=0 will give the following approximate
k+1 ∂ x2 solution:
∞
where the Uk (x), is the transform function of the
∑ Uk (x)t k
⌢
u (x,t) =
t−dimensional spectrum. k=0

U0 (x) = sinh(x). (26) = U0 (x) +U1 (x)t +U2 (x)t 2 +U3 (x)t 3 + ...
t2 t3 t4
Now, substitute Eq. (26) into Eq. (25) to obtain the = cosh(x) − 1 + t − + − + ...
2 6 12
following:
t2 t3 t4
 
= cosh(x) − 1 − t + − + + ...
U1 (x) = i sinh(x) 2 6 24
= cosh(x) − e−t .
U2 (x) = 21 sinh(x)
This is an exact solution of Eq. (27).
−i
U3 (x) = 6 sinh(x).
We continue in this manner and after a few iterations, the
differential inverse transform of {Uk (x)}∞
4 Tables of Numerical Calculations
k=0 will provide
us with the following approximate solution:
In this section, we shall illustrate the accuracy and
∞
∑ Uk (x)t efficiency of the RDTM. For this purpose, we can
⌢ k
u (x,t) =
k=0 evaluate the approximate solution using the 5th-order
= U0 (x) +U1 (x)t +U2 (x)t 2 + ....
approximation. Table 2 shows the exact solution, the
approximate solution and the absolute error obtained by
t2 it 3 the RDTM. We must emphasize here only five iterations
= sinh(x) + it sinh(x) − sinh(x) − sinh(x) + ...
2 6 was used for different values of x and t, specifically,
t 2 it 3
 
= 1 + it − − + ... sinh(x)
x = −5, −3, 3, 5 and t = 0.002, 0.004, 0.006, 0.01.
2 6
= eit sinh(x). Table 2: Comparison of the absolute error of the solutions of the STO equation by
RDTM
This is the exact solution of Eq. (22). x
−5
t
.002
Exact Solution
−0.4933204320
RDTM Solution
−0.4933204320
Abs-error-RDTM, n=5
5.55111512E −17
.004 −0.4933336888 −0.4933336888 0
.006 −0.4933469195 −0.4933469195 0
.01 −0.4933733027 −0.4933733027 5.55111512E −17
−3 −0.4526643984 −0.4526643984
3.3 Telegraph equation .002
.004 −0.4527545066 −0.4527545066
0
5.55111512E −17
.006 −0.4528444519 −0.4528444519 5.55111512E −17
.01 −0.4530238543 −0.4530238543 5.55111512E −17
We consider the homogeneous Telegraph equation of the 3 .002 0.4524836916 0.4524836916 5.55111512E −17
form: .004 0.4523930926 0.4523930926 0
.006 0.4523023296 0.4523023296 5.55111512E −17
5.55111512E −17
uxx = utt − 2ut − u, (27) 5
.01
.002
0.4521203101
0.4932938398
0.4521203101
0.4932938398 0
.004 0.4932805043 0.4932805043 0
subject to the initial condition .006 0.4932671424 0.4932671424 0
.01 0.49324033948 0.49324033948 5.55111512E −17

u(x, 0) = cosh(x) − 1; ut (x, 0) = 1, (28)

where the exact solution

u(x,t) = cosh(x) − e−t . (29)

5 Conclusion
Now, we apply the RDTM to Eq. (27) and Eq. (28) we get
In this paper, we applied the Reduced Differential
∂ 2 Uk (x)
!
1
Uk+2 (x) =
(k + 2)(k + 1) ∂ x2
− 2(k + 1)Uk+1 (x) −Uk (x) (30) Transform Method (RDTM) to all three physical models,

c 2014 NSP
Natural Sciences Publishing Cor.
Appl. Math. Inf. Sci. 8, No. 5, 2171-2176 (2014) / www.naturalspublishing.com/Journals.asp 2175

namely, the Sharma Tasso Olver (STO) equation, the [11] A. M. Wazwaz, Partial Differential Equations: Methods and
Schrodinger equation and the Telegraph equation. We Applications, Balkema, Leiden, (2002).
successfully found approximate solution for the STO and [12] B. Soltanalizadeh, Application of Differential
the results we obtained in example (3.1) were in excellent Transformation Method for Numerical Analysis of
agreement with the exact solution. Also, we found exact Kawahara Equation, Australian Journal of Basic and
solutions to the Schrodinger equation and the Telegraph Applied Sciences, 12, 490-495 (2011).
equation. The RDTM introduces a significant [13] B. bi, M. Bayram, Approximate Solutions for Some
Nonlinear Evolutions Equations By Using The Reduced
improvement in the fields over existing techniques
Differential Transform Method, International Journal of
because it takes less calculations and the number of
Applied Mathematical Research, 3, 288-302 (2012).
iteration is less compared by other methods. My goal in [14] M. Rawashdeh, Improved Approximate Solutions for
the future is to apply the RDTM to fractional nonlinear Nonlinear Evolutions Equations in Mathematical Physics
PDEs that arises in other areas of science such as Biology, Using the RDTM, Journal of Applied Mathematics and
Medicine and Engineering. There is no existing method in Bionorfomatics, 3, 1-14 (2013).
the literature that can give exact solution to fractional [15] M. Rawashdeh, Using the Reduced Differential Transform
PDEs, so finding approximate solutions is very important. Method to Solve Nonlinear PDEs Arises in Biology and
Computations of this paper have been carried out using Physics, World Applied Sciences Journal, 8, 1037-1043
the computer package of Mathematica 7. (2013).
[16] R. Abazari, B. Soltanalizadeh, Reduced Differential
Transform Method and Its Application on Kawahara
6 Acknowledgement Equations, Thai Journal of Mathematics, 11, 199-216
(2013).
The author would like to thank the Editor and the [17] S. Wang, X. Tang and S. Lou, Soliton fission and fusion:
anonymous referees for their comments and suggestions Burgers equation and Sharma Tasso Olver equation, Chaos,
on improving this paper. This research was supported by Solitons and Fractals, International Journal of Applied
the Deanship of Research at Jordan University of Science Mathematical Research, 21, 231-2239 (2004).
and Technology No. (87/2013).

Mahmoud Rawashdeh joined

References the Department of Mathematics
and Statistics at Jordan University
[1] Y. Keskin, G. Oturan, Reduced Differential Transform of Science and Technology
Method for Partial Differential Equations, International (Jordan) in 2009. Prior to coming
Journal of Nonlinear Sciences and Numerical Simulation, to the Faculty of Science and
6, 741-749 (2009). Arts at (JUST), he was a tenured
[2] Y. Keskin, G. Oturan, Reduced Differential Transform assistant professor at the University
Method for fractional partial differential equa- tions, of Findlay (USA) (2006–2009).
Nonlinear Science Letters A, 2, 61-72 (2010).
He received his undergraduate
[3] Y. Keskin, G. Oturan, Application of Reduced Differential
degree in Mathematics from
Transformation Method for Solving Gas Dynamics
Yarmouk University (Jordan) in 1989. He received his M.A in
Equation, Int. J. Contemp. Math. Sciences, 22, 1091-1096
(2010). Mathematics from City University of New York, New York,
[4] Y. Keskin, Ph.D Thesis, University of Secuk (in Turkish), USA in June 1997 and his Ph.D. in Applied Mathematics from
(2010). The University of Toledo, Ohio, USA in May 2006. He was
[5] G. Adomian, Solving frontier problems of physics: the recognized for excellence in teaching in the ”Owens Exchange
decomposition method, Kluwer Acad. Publ, (1994). newsletter” of Owens Community College, September 2004. He
[6] G. Adomian, A new approach to nonlinear partial also served on the College of Science Committee at UF
differential equations, J. Math. Anal. Appl., 102, 420-434 (2006–2009),the Institutional Review Board (IRB) at UF
(1984). (2007–2009) and the Committee on Committee at UF
[7] M. Alquran, Applying Differential Transform Method (2008-2009). Moreover, Dr. Rawashdeh was a co-chair of the
to Nonlinear Partial Differential Equations: A Modified Hiring Committee for Two new Mathematics positions (UF,
approach, Applications and Applied Mathematics: An 2007–2008). His research interest include topics in the areas of
International Journal, 7, 155-163 (2012). Applied Mathematics, such as; Lie Algebra, Mathematical
[8] A. M. Wazwaz, Partial Differential Equations and Solitary
Physics and Functional Analysis (Approximation Theory).
Waves Theory, Springer–Verlag, Heidelberg, (2009).
His research involves using numerical methods to obtain
[9] A. M. Wazwaz, A sine–cosine method for handling
approximate and exact solutions to partial differential equations
nonlinear wave equations, Math. Comput. Modelling, 40,
499-508 (2004). arising from nonlinear PDEs problems in engineering and
[10] A. M. Wazwaz, New solitons and kinks solutions to the Physics. He has published research articles in a well-recognized
Sharma Tasso Olver equation, Applied Mathematics and international journals of mathematical and engineering sciences.
Computation, 188, 1205-1213 (2007). He is a referee in a highly respected mathematical journals.

c 2014 NSP
Natural Sciences Publishing Cor.
2176 M. Rawashdeh, N. Obeidat: On Finding Exact and Approximate Solutions to Some...

Nazek Obeidat received

her undergraduate degree
in Mathematics from
the University of Findlay
(USA) in 2008, graduating
with summa cum laude. She
received her M.Sc in Applied
Mathematics from Jordan
University of Science and
Technology in 2012. She was
recognized for excellent academic achievement and made
the Dean’s list 2004 and 2005 at the University of Toledo
and recognized for excellent academic achievement and
made the Dean’s list 2006, 2007, and 2008 at the
University of Findlay. She already has one paper
published (in press) in a good international journal and
another paper submitted. Her research involve solving
linear and nonlinear PDEs in the area of Physics and
Engineering. Her goal in the near future to pursue a Ph.D
degree in Mathematics.

c 2014 NSP
Natural Sciences Publishing Cor.