SlideShare a Scribd company logo
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

A Class of A-Stable Order Four and Six Linear Multistep
Methods for Stiff Initial Value Problems
Kumleng G.M1, Longwap S1, Adee S. O2
1. Department of Mathematics University of Jos, P.M.B. 2084, Plateau State, Nigeria
2. Department of Mathematics, Fed. University of Tech. Yola, Nigeria
* E-mail of the corresponding author: kumleng_g@yahoo.com
Abstract
A new three and five step block linear methods based on the Adams family for the direct solution of stiff initial
value problems (IVPs) are proposed. The main methods together with the additional methods which constitute
the block methods are derived via interpolation and collocation procedures. These methods are of uniform order
four and six for the three and five step methods respectively. The stability analysis of the two methods indicates
that the methods are A–stable, consistent and zero stable. Numerical results obtained using the proposed new
block methods show that they are attractive for the solutions of stiff problems and compete favorably with the
well-known Matlab stiff ODE solver ODE23S.
Keywords: Linear multistep methods, initial value problems, interpolation and collocation.
1. Introduction
In this paper, we shall be concerned with the approximate numerical integration of the stiff initial value problem

y ' = f ( x, y ), y ( x0 ) = y0 , x ∈ [ a, b], y ∈ R

(1)
Numerical analysts have focused most of their works on the development of more efficient and accurate methods
for the solution of stiff problems, and as such many methods have been proposed. Notable among these methods
is the Backward Differentiation formulae (BDF), because of its wide region of absolute stability. Several
researchers such as Kim (2010), Chollom et. al. (2012), Kumleng et.al (2011, 2012) have proposed methods for
the numerical solution of (1) which were shown to be very efficient.
In this paper, we describe the construction of a new three and five step methods based on the Adams
methods similar to the generalized Adams methods of Brugnano and Trigiante (1998). These methods are
constructed using the interpolation concept where the continuous interpolants provide the block methods through
evaluation at some grid points. This approach produce two self- starting new A – stable block methods which
provide the solutions of stiff initial value problems on non-overlapping intervals.
2. The New Method
In this section, the new A – stable block methods shall be constructed based on the continuous finite difference
approximation approach using the interpolation and collocation criteria described by Lie and Norsett (1981)
called multistep collocation (MC) and block multistep methods by Onumanyi et al. (1994).
We define based on the interpolation and collocation methods the continuous form of the k- step new
method as
m

yn + v − α v −1 ( x) yn+ v −1 = h∑ β j ( x) f n+ j
j =0

where

α v−1 and β j ( x)

(2)

are the continuous coefficients of the method, m is the number of distinct collocation

points, h is the step size and

v=

k −1
2

, k = 3,5, 7...

From Onumanyi et. al (1994), we obtain our matrices D and C = D

1 xn+ v −1

1
0
D = 0
1

M
0
1


2
xn+ v −1 L

2 x0
2 x1

L
L

2 xm −1

L
L

m
xn + v −1 
m 
mx0 −1 
mx1m −1 

M 
m −1
mxm −1 


−1

as

(3)

and
1
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

 α v −1,1
α
v −1,2
C=
 M

α v −1,m +1

hβ 0,1
hβ 0,2
hβ 0,m +1

www.iiste.org

hβ m−1,1 
hβ m −1,2 


M

L hβ m−1,m+1 
L
L

(4)

respectively.
2.1 Derivation of the Three Step New Block Method
In this case, k = 3, v=1, t = 1 and m = 4 and its continuous form expressed in the form of (2) is;

y ( x ) = α1 ( x ) yn + h ( β 0 ( x) f n + β1 ( x ) f n +1 + β 2 ( x) f n + 2 + β3 ( x ) f n + 3 )

(5)

The matrix D, in (3) becomes

 1 xn

0 1
D = 0 1

0 1
0 1


2
xn

3
xn

2 xn
2 xn+1

2
3 xn
2
3 xn +1

2 xn + 2
2 xn +3

2
3 xn + 2
2
3 xn +3

4
xn 
3 
4 xn 
3
4 xn +1 

3
4 xn + 2 
3 
4 xn +3 

(6)

Using the Maple software, the inverse of the matrix in (6) is obtained and this yields the elements of the matrix C.
The elements of the matrix C substituted into (5) yields the continuous formulation of the method as:

(
+ (−

)

λ
y( xn + λ ) = yn + −λ − 11λh + 3λh2 − 24 h3 f n +
12
2

3λ
4h

2

3

4

)

+ 2λ2 − 8λh3 f n+ 2 +
3h
3

Evaluating (7) at the following points λ = h,
constitute the new three step block method.

4

(

λ

2

6h

(

3λ 2
2h

)

− 5λ2 + 8λh3 f n +1
6h
3

4

)

λ
− 6λh2 + 24 h3 f n +3
3

4

λ = 2h, λ = 3h yields the following discrete methods which

( 9 f n + 19 f n +1 − 5 f n + 2 + f n +3 )
yn + 2 − yn = h ( f n + 4 f n +1 + f n + 2 )
3
h
y n + 3 − y n = 8 ( 3 f n + 9 f n +1 + 9 f n + 2 + 3 f n + 3 )
yn +1 − yn =

(7)

h
24

(8)

The main new discrete scheme for the three step method is

yn +1 − yn =

h
24

( 9 f n + 19 f n +1 − 5 f n +2 + f n +3 )

(9)
This new method is consistent since its order is 4, it is also zero-stable, above all, it is A – stable as can be seen
in figure 1. The new three step discrete methods that constitute the block method (8) have the following orders
and error constants as can be seen in Table 1.
2.2 Derivation of the Five –Step New Block Method
In this case, k = 5,
t = 1 and m = 6 and its continuous form expressed in the form of (2) is;

y ( x ) = α1 ( x ) yn +1 + h ( β 0 ( x ) f n + β1 ( x ) f n +1 + β 2 ( x ) f n + 2 + β 3 ( x ) f n + 3 + β 4 ( x ) f n + 4 + β 5 ( x ) f n + 5 )

Similarly, we generate the continuous formulation of the new five step method as;

2

(10)
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

(
+ (− h + −
+
−
+( h− +
−
+
+(− h + −
+
−
+( h − +
−
+
+(− h + −
+
−

)

95
λ
λ
y( xn + λ ) = yn+1 + − 288 h + λ − 137λh + 5λ2 − 17λ3 + 40h4 − 720h5 f n
120
8h
96 h
2

5λ 2
2h

1427
1440

5λ 2
2h

133
240

173
1440

77 λ 3
36 h2

5λ
8h

3
160

71λ 4
96 h3

13λ 5
60 h4

13λ 3
6 h2

49λ 4
48 h3

5 h4

3

λ2

5λ 3
36 h2

10 h

41λ
96 h3

4

7λ4
90 h3

5

6

6

6

6

11λ
120 h4

λ
− 144 h5

λ5

λ
+ 720h5

5

60 h4

)
)f
)f
)f

λ
+ 144h5 f n+2
λ
+ 720h5

λ5

)

λ
+ 144h5 fn +1

7 λ5
60 h4

59λ 4
48h3

61λ
72 h2

2

4

107 λ 3
36 h2

5λ 2
3h

241
720

3

6

6

n +3

n +4
n +5

(11)

Evaluating (11) at the following points λ = 0, λ = 2 h, λ = 3h, λ = 4 h, λ = 5h yields the following
discrete methods which constitute the new five step block method.
h
yn + 1 - yn = 1440 ( 475 fn + 1427 fn + 1 - 798 fn + 2 + 482 fn + 3 -173fn + 4 + 27 fn + 5 )
h
yn + 2 - yn+1 = 1440 ( -27 fn + 637 fn + 1 +1022 fn + 2 -258 fn + 3 +77 fn + 4 -11 fn + 5 )
h
yn + 3 - yn+1 = 90 ( - fn + 34 fn + 1 +114 fn + 2 +34 fn + 3 - fn + 4 )
h
yn + 4 - yn+1 = 160 ( -3fn + 69 fn + 1 +174 fn + 2 +174 fn + 3 +69 fn + 4 -3 fn + 5 )
h
yn + 5 - yn+1 = 45 (14 fn + 1 +64 fn + 2 +24 fn + 3 +64 fn + 4 +14 fn + 5 )

(12)

The main new discrete scheme for the five step method is
h
yn + 2 - yn +1 = 1440 ( −27 f n + 637 f n + 1 + 1022 f n + 2 − 258 f n + 3 + 77 f n + 4 − 11 f n + 5 )

(13)
This new method is consistent since its order is 6, its also zero-stable, above all, it is A – stable as can be seen in
figure 1.
The new five step discrete methods (12) have the following orders and error constants as shown in Table 1.
3. Analysis of the New Methods
In this section, we consider the analysis of the newly constructed methods. Their convergence is determined and
their regions of absolute stability are plotted.
3.1 Convergence
The convergence of the new block methods is determined using the approach by Fatunla (1991) and Chollom et.
al (2007) for linear multistep methods, where the block methods are represented in a single block, r point
multistep method of the form
k

k

i =1

i =0

A(0) ym +1 = ∑ A(i ) ym+1 + h∑ B (i ) f m −1
(14)

where h is a fixed mesh size within a block, A , B , i = 0,1, 2,...k are
i

ym , ym −1 and ym +1

i

r × r identity matrix while

are vectors of numerical estimates.

Definition 1: A block method is zero stable provided the roots

polynomial ρ ( R ) specified as



k

Rij = 1, 2,...k

of the first characteristic



ρ ( R) = det ∑ A(i ) R k −i  = 0
 i =0

satisfies

Rj ≤ 1



(15)

, the multiplicity must not exceed two. Fatunla (1994)
3
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

The block method (8) expressed in the form of (14) gives

 19
 1 0 0   yn +1   0 0 1   yn − 2 
24
 4
 




 0 1 0   yn + 2  =  0 0 1   yn −1  + h  3
 0 0 1   y   0 0 1  y 
 9

  n +3  
 n 
 8

−5
24

  f n +1   0 0
 

0   fn+2  +  0 0
3 
 0 0
8   f n +3 


1
24

1
3
9
8

9
24
1
3
3
8

  f n−2  


  f n −1  


  fn 


(16)

where
−5
9
1
 19 24 24 
 0 0 24 
1 0 0
 0 0 1
24








A(0) =  0 1 0  , A(1) =  0 0 1 , B (0) =  4 1 0  , B (1) =  0 0 1 
3
3
3
0 0 1
 0 0 1
3 
9 9 3





8 8 8
0 0 8 
(0)
(1)
Substituting A and A into (15) gives the characteristic polynomial of the block method (8) as
ρ (λ ) = det(λ A(0) − A(1) )

  1 0 0   0 0 1 
 
 

= det λ  0 1 0  −  0 0 1 
 

  0 0 1   0 0 1 
 

 λ 0 −1 
= det  0 λ −1 


 0 0 λ − 1


= λ 2 (λ − 1) = 0
λ = 1, λ2 = λ3 = 0 . The block method (8) by definition1 is zero stable and by Henrici (1962), the
Therefore, 1
block method is convergent since it is also consistent.
Similarly, the block method (12) expressed in the form of (14) gives

1

 −1
 −1

 −1
 −1


0 0 0 0   yn +1   0
 

1 0 0 0   yn + 2   0
0 1 0 0   yn + 3  =  0
 

0 0 1 0   yn + 4   0


0 0 0 1   yn + 5   0



 1427
1440
 637
 1440
h  34
90
 69
 160
 14
 45

−798
1440
1022
1440
114
90
174
160
64
45

482
1440
−258
1440
34
90
174
160
24
45

−173
1440
77
1440
−1
90
69
160
64
45


−11 
1440 
0 

−3
160 
14  
45  

27
1440

0 0 0 1   yn − 4 


0 0 0 0   yn − 3 
0 0 0 0   yn − 2  +


0 0 0 0   yn −1 


0 0 0 0   yn 

f n +1   0
 
f n+ 2   0
f n +3  +  0
 
f n+ 4   0

f n +5   0


0 0 0
0 0 0
0 0 0
0 0 0
0 0 0




−1  
90

−3
160  

0 

475
1440
−27
1440

f n− 4  

f n −3  
f n− 2  

f n −1  

fn 

(17)

where

4
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

A(0)

1

 −1
=  −1

 −1
 −1


B (0)

 1427
1440
 637
 1440
=  34
90
 69
 160
 14
 45

Substituting

www.iiste.org

0 0 0 0
0 0


1 0 0 0
0 0
(1)
0 1 0 0, A =  0 0


0 0 1 0
0 0


0 0 0 1
0 0
−798
1440
1022
1440
114
90
174
160
64
45

482
1440
−258
1440
34
90
174
160
24
45

−173
1440
77
1440
−1
90
69
160
64
45

0 0 1

0 0 0
0 0 0

0 0 0

0 0 0


0 0


−11
1440 
0 0
(1)
0 , B = 0 0


−3
160 
0 0
0 0
14 

45 

27
1440

0 0
0 0
0 0
0 0
0 0




−1 
90

−3
160 

0 

475
1440
−27
1440

A(0) and A(1) into (15) gives the characteristic polynomial of the block method (12) as
ρ (λ ) = det(λ A(0) − A(1) )
 1
 
  −1
= det λ  −1
 
  −1
  −1
 
λ
 −λ

= det  −λ

 −λ
 −λ


0 0 0 0 0
 
1 0 0 0 0
0 1 0 0 − 0
 
0 0 1 0 0

0 0 0 1 0


0

0

0

λ

0

0

0 λ 0
0 0 λ
0

0

0

0 0 0 1 

0 0 0 0 
0 0 0 0 

0 0 0 0 

0 0 0 0 


−1
0

0

0
λ


= λ5 − λ4 = 0
λ = 1, λ2 = λ3 = λ4 = λ5 = 0 . The block method (12) by definition1 is zero stable and by Henrici
Therefore, 1
(1962), it is convergent since it is also consistent.
3.2 Regions of Absolute Stability of the Methods
The absolute stability regions of the newly constructed block methods (8) and (12) are plotted using Chollom
(2005) by reformulating the methods as general linear methods. The regions of absolute stability of the methods
of the main discrete schemes and the new block methods are as shown in the figures 1 and 2 below. These
absolute stability regions are all A –stable since they consist of the set of points in the complex plane outside the
enclosed figures.
4 Numerical Examples
We report here a few numerical examples on some stiff problems taken from the literature. For comparisons, we
also report the performance of the new block methods and the well-known Matlab stiff ODE solver ODE23S on
the same problems.

5
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

Problem 1: We consider the Robertson’s problem

y '1 = −0.04 y1 + 10000 y2 y3

y1 (0) = 1

2
y '2 = 0.04 y1 − 10000 y2 y3 − 30000000 y2

y2 (0) = 0

2
y '3 = 30000000 y2

y3 (0) = 0

0 ≤ x ≤ 70, h = 0.1
Problem 2: We consider the Van der Pol’s equation

y '1 = y2

y1 (0) = 2

y '2 = - y1 + µ y2 (1- y12 ) ,

y2 (0) = 0

0 ≤ x ≤ 70, h = 0.01, µ = 10,
5. Conclusion
The new block methods were applied to two well-known stiff problems from the literature, the numerical results
suggest that the proposed new block methods (8) and (12) are suitable for solving stiff problems and perform
competitively with the well-known ODE23s. This success is achieved because of the good stability properties of
the proposed new block methods.
References
Brugnano, L. and D, Trigiante. (1998). Solving Differential Problems by Multistep Initial and Boundary Value
Methods. Gordon and Breach Science Publishers, Amsterdam.
Chollom, J.P., (2005). The construction of Block Hybrid Adam Moulton Methods with link to Two Step RungeKutta Methods. (Ph.D) Thesis University of Jos.
Chollom, J.P., Ndam, J.N. and Kumleng, G.M. (2007). On Some Properties of the Block Linear Multistep
Methods. Science World Journal, 2(3), 11 – 17.
Chollom, J.P., Olatunbasun,I, O. and Omagu, S. (2012). A Class of A-Stable Block Explicit Methods for the
Solutions of ODEs Research Journal of Mathematics and Statistics, 4(2): 52-56.
Fatunla, S.O. (1991). Parallel Method for Second Order Differential Equations. Proceedings of the National
Conference on Computational Mathematics, University of Ibadan Press, pp:87-99.
Fatunla, S.O. (1994). Block Method for Second Order Differential Equations. International Journal of
Computer Mathematics. 41,55 – 63.
Henrici, P. (1962). Discrete Variable Methods in Ordinary Differential Equations. John Wiley, New York, p407
Kim, P. (2010). An Explicit Type Stable Method for Solving Stiff Initial Value Problems. Presentation at a mini
workshop at Knu. Republic of Korea, Retrieved from: http://webbuild.knu.ac.kr/~skim/conf_math2/kimps.pdf.
Kumleng,G.M. and Skwame,Y. (2011) A New A-Stable Method for the Solution of Stiff Initial Value
Problems. International Journal of Numerical Maths.6 (2), 360 -373.
Kumelng,G.M., Sirisena, U.W. and Chollom, J.P.(2012) Construction of a Class of Block Hybrid Implicit
Multistep Methods for the Solution of Stiff Ordinary Differential Equations. Nigerian Journal of Pure and
Applied Sciences 5.
Lie, I. and Norsett, S.P. (1989). Super convergence for Multistep Collocation. Mathematics of Computation.
52(185), 65 – 79.
Onumanyi, P., Awoyemi, D.O., Jator, S.N. and Sirisena, U.W. (1994).New Linear Multistep Methods with
Continuous Coefficients for First Order Initial Value Problems. Journal of the Nigerian Mathematical
Society .13, 37 – 51.

6
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

2.5
2
1.5
1

I m (z )

0.5
0
k=3

k=5

-0.5
-1
-1.5
-2
-2.5

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Re(z)

Figure2. Absolute Stability Regions of the new Three and Five step discrete methods
2.5
2
1.5
1

I m (z )

0.5
0
k=5

k=3

-0.5
-1
-1.5
-2
-2.5
-0.5

0

0.5

1
Re(z)

1.5

2

2.5

Figure1. Absolute Stability Regions of the Three and Five step New block methods

The absolute stability regions consists of the set of points in the complex plane outside the enclosed figures.
Therefore, both the discrete and block methods are all A - stable since the left –half complex plane is contained
in S., where S = { z ∈ C : R ( z ) ≤ 1} .

7
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

1

y1 Num
y2 Num
y3 Num
y1 ODE23s
y2 ODE23s
y3 ODE23s

0.9
0.8
0.7
S o lu t io n (y )

0.6
0.5
0.4
0.3
0.2
0.1
0

0

10

20

30

40

50

60

70

Time(t)

Figure 1. Solution to Problem 1 using the Three Step Block Method and ODE23S
y1 Num
y2 Num
y3 Num
y1 ODE23s
y2 ODE23s
y3 ODE23s

1
0.9
0.8

S o lu tio n (y )

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0

0

10

20

30

40

50

60

70

Time(t)

Figure 2. Solution to Problem 1 using the Five Step Block Method and ODE23S

15

y1 Num
y2 Num
y1 ODE23s
y2 ODE23s

10

S o lu t io n ( y )

5

0

-5

-10

-15

0

10

20

30

40

50

60

70

Time(t)

Figure 3. Solution to Problem 2 using the Three Step Block Method and ODE23S

8
Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.11, 2013

www.iiste.org

15
y1 Num
y2 Num
y1 ODE23s
y2 ODE23s

10

S o lu t io n ( y )

5

0

-5

-10

-15

0

10

20

30

40

50

60

70

Time(t)

Figure 4. Solution to Problem 2 using the Five Step Block Method and ODE23S
Table 1: Order and Error Constants of the new three step method
Method

y n +1

Order
4

Error constant

−

4

y n+2

1
− 90
3
− 80

4

y n+3

19
720

Table 2: Order and Error Constant of the new five step method
Method

y n +1
y n+2
y n+3

Order
6

Error constant
863
− 60480
271
60480

6

1
756

6

y n+4
y n+5

13
2240

6

6

8
− 945

9
This academic article was published by The International Institute for Science,
Technology and Education (IISTE). The IISTE is a pioneer in the Open Access
Publishing service based in the U.S. and Europe. The aim of the institute is
Accelerating Global Knowledge Sharing.
More information about the publisher can be found in the IISTE’s homepage:
http://www.iiste.org
CALL FOR JOURNAL PAPERS
The IISTE is currently hosting more than 30 peer-reviewed academic journals and
collaborating with academic institutions around the world. There’s no deadline for
submission. Prospective authors of IISTE journals can find the submission
instruction on the following page: http://www.iiste.org/journals/
The IISTE
editorial team promises to the review and publish all the qualified submissions in a
fast manner. All the journals articles are available online to the readers all over the
world without financial, legal, or technical barriers other than those inseparable from
gaining access to the internet itself. Printed version of the journals is also available
upon request of readers and authors.
MORE RESOURCES
Book publication information: http://www.iiste.org/book/
Recent conferences: http://www.iiste.org/conference/
IISTE Knowledge Sharing Partners
EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open
Archives Harvester, Bielefeld Academic Search Engine, Elektronische
Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial
Library , NewJour, Google Scholar

More Related Content

PDF
11.a family of implicit higher order methods for the numerical integration of...
PDF
Implicit two step adam moulton hybrid block method with two off step points f...
PDF
Sixth order hybrid block method for the numerical solution of first order ini...
PDF
Derivation and Application of Six-Point Linear Multistep Numerical Method for...
PDF
Bit sat 2008 questions with solutions
PDF
Uniform Order Legendre Approach for Continuous Hybrid Block Methods for the S...
PDF
Introductory Mathematical Analysis for Business Economics International 13th ...
PDF
E029024030
11.a family of implicit higher order methods for the numerical integration of...
Implicit two step adam moulton hybrid block method with two off step points f...
Sixth order hybrid block method for the numerical solution of first order ini...
Derivation and Application of Six-Point Linear Multistep Numerical Method for...
Bit sat 2008 questions with solutions
Uniform Order Legendre Approach for Continuous Hybrid Block Methods for the S...
Introductory Mathematical Analysis for Business Economics International 13th ...
E029024030

What's hot (16)

PDF
Coueete project
PDF
Chapter 3 linear equations
DOCX
.Chapter7&8.
PDF
Research Inventy : International Journal of Engineering and Science
PDF
Accurate Numerical Method for Singular Initial-Value Problems
PPT
Chapter2 functionsandgraphs-151003144959-lva1-app6891
PDF
D0366025029
PDF
Module 1 linear functions
PDF
PDF
C142530
PPT
Gauss Jordan
PPT
Introductory maths analysis chapter 02 official
PDF
Engineering Mathematics 2 questions & answers
PDF
SUCCESSIVE LINEARIZATION SOLUTION OF A BOUNDARY LAYER CONVECTIVE HEAT TRANSFE...
PDF
Mathematics xii paper 13 with answer with value vased questions
Coueete project
Chapter 3 linear equations
.Chapter7&8.
Research Inventy : International Journal of Engineering and Science
Accurate Numerical Method for Singular Initial-Value Problems
Chapter2 functionsandgraphs-151003144959-lva1-app6891
D0366025029
Module 1 linear functions
C142530
Gauss Jordan
Introductory maths analysis chapter 02 official
Engineering Mathematics 2 questions & answers
SUCCESSIVE LINEARIZATION SOLUTION OF A BOUNDARY LAYER CONVECTIVE HEAT TRANSFE...
Mathematics xii paper 13 with answer with value vased questions
Ad

Similar to A class of a stable order four and six linear multistep methods for stiff initial value problems (20)

PDF
F0333034038
PDF
A (  )-Stable Order Ten Second Derivative Block Multistep Method for Stiff I...
PDF
A Class of Continuous Implicit Seventh-eight method for solving y’ = f(x, y) ...
PDF
paper publication
PDF
International Journal of Mathematics and Statistics Invention (IJMSI)
PDF
International Journal of Mathematics and Statistics Invention (IJMSI)
PDF
General linear methods for ordinary differential equations
PDF
Chebyshev Collocation Approach for a Continuous Formulation of Implicit Hybri...
PDF
Derivation and Application of Multistep Methods to a Class of First-order Ord...
PDF
Hybrid Block Method for the Solution of First Order Initial Value Problems of...
PDF
New Two-Step Method with Fifth-Order Convergence for Solving Nonlinear Equations
PDF
Ma2002 1.14 rm
PDF
04_AJMS_299_21.pdf
PDF
General linear methods
PDF
05_AJMS_332_21.pdf
PDF
Ma2002 1.13 rm
PDF
Block Hybrid Method for the Solution of General Second Order Ordinary Differe...
PDF
Order and stability for general linear methods
PDF
Applications Of MATLAB Ordinary Differential Equations (ODE
PDF
Numerical Method for UOG mech stu prd by Abdrehman Ahmed
F0333034038
A (  )-Stable Order Ten Second Derivative Block Multistep Method for Stiff I...
A Class of Continuous Implicit Seventh-eight method for solving y’ = f(x, y) ...
paper publication
International Journal of Mathematics and Statistics Invention (IJMSI)
International Journal of Mathematics and Statistics Invention (IJMSI)
General linear methods for ordinary differential equations
Chebyshev Collocation Approach for a Continuous Formulation of Implicit Hybri...
Derivation and Application of Multistep Methods to a Class of First-order Ord...
Hybrid Block Method for the Solution of First Order Initial Value Problems of...
New Two-Step Method with Fifth-Order Convergence for Solving Nonlinear Equations
Ma2002 1.14 rm
04_AJMS_299_21.pdf
General linear methods
05_AJMS_332_21.pdf
Ma2002 1.13 rm
Block Hybrid Method for the Solution of General Second Order Ordinary Differe...
Order and stability for general linear methods
Applications Of MATLAB Ordinary Differential Equations (ODE
Numerical Method for UOG mech stu prd by Abdrehman Ahmed
Ad

More from Alexander Decker (20)

PDF
Abnormalities of hormones and inflammatory cytokines in women affected with p...
PDF
A validation of the adverse childhood experiences scale in
PDF
A usability evaluation framework for b2 c e commerce websites
PDF
A universal model for managing the marketing executives in nigerian banks
PDF
A unique common fixed point theorems in generalized d
PDF
A trends of salmonella and antibiotic resistance
PDF
A transformational generative approach towards understanding al-istifham
PDF
A time series analysis of the determinants of savings in namibia
PDF
A therapy for physical and mental fitness of school children
PDF
A theory of efficiency for managing the marketing executives in nigerian banks
PDF
A systematic evaluation of link budget for
PDF
A synthetic review of contraceptive supplies in punjab
PDF
A synthesis of taylor’s and fayol’s management approaches for managing market...
PDF
A survey paper on sequence pattern mining with incremental
PDF
A survey on live virtual machine migrations and its techniques
PDF
A survey on data mining and analysis in hadoop and mongo db
PDF
A survey on challenges to the media cloud
PDF
A survey of provenance leveraged
PDF
A survey of private equity investments in kenya
PDF
A study to measures the financial health of
Abnormalities of hormones and inflammatory cytokines in women affected with p...
A validation of the adverse childhood experiences scale in
A usability evaluation framework for b2 c e commerce websites
A universal model for managing the marketing executives in nigerian banks
A unique common fixed point theorems in generalized d
A trends of salmonella and antibiotic resistance
A transformational generative approach towards understanding al-istifham
A time series analysis of the determinants of savings in namibia
A therapy for physical and mental fitness of school children
A theory of efficiency for managing the marketing executives in nigerian banks
A systematic evaluation of link budget for
A synthetic review of contraceptive supplies in punjab
A synthesis of taylor’s and fayol’s management approaches for managing market...
A survey paper on sequence pattern mining with incremental
A survey on live virtual machine migrations and its techniques
A survey on data mining and analysis in hadoop and mongo db
A survey on challenges to the media cloud
A survey of provenance leveraged
A survey of private equity investments in kenya
A study to measures the financial health of

Recently uploaded (20)

PDF
MIND Revenue Release Quarter 2 2025 Press Release
PDF
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
PDF
Electronic commerce courselecture one. Pdf
PDF
A comparative analysis of optical character recognition models for extracting...
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
PDF
Review of recent advances in non-invasive hemoglobin estimation
PDF
Diabetes mellitus diagnosis method based random forest with bat algorithm
PPTX
Machine Learning_overview_presentation.pptx
PDF
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
PDF
Reach Out and Touch Someone: Haptics and Empathic Computing
PPTX
Big Data Technologies - Introduction.pptx
PDF
NewMind AI Weekly Chronicles - August'25-Week II
PDF
Chapter 3 Spatial Domain Image Processing.pdf
PDF
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
PPTX
MYSQL Presentation for SQL database connectivity
PPTX
20250228 LYD VKU AI Blended-Learning.pptx
PDF
Encapsulation theory and applications.pdf
PDF
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
PPT
“AI and Expert System Decision Support & Business Intelligence Systems”
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
MIND Revenue Release Quarter 2 2025 Press Release
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
Electronic commerce courselecture one. Pdf
A comparative analysis of optical character recognition models for extracting...
Dropbox Q2 2025 Financial Results & Investor Presentation
Review of recent advances in non-invasive hemoglobin estimation
Diabetes mellitus diagnosis method based random forest with bat algorithm
Machine Learning_overview_presentation.pptx
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Reach Out and Touch Someone: Haptics and Empathic Computing
Big Data Technologies - Introduction.pptx
NewMind AI Weekly Chronicles - August'25-Week II
Chapter 3 Spatial Domain Image Processing.pdf
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
MYSQL Presentation for SQL database connectivity
20250228 LYD VKU AI Blended-Learning.pptx
Encapsulation theory and applications.pdf
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
“AI and Expert System Decision Support & Business Intelligence Systems”
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...

A class of a stable order four and six linear multistep methods for stiff initial value problems

  • 1. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org A Class of A-Stable Order Four and Six Linear Multistep Methods for Stiff Initial Value Problems Kumleng G.M1, Longwap S1, Adee S. O2 1. Department of Mathematics University of Jos, P.M.B. 2084, Plateau State, Nigeria 2. Department of Mathematics, Fed. University of Tech. Yola, Nigeria * E-mail of the corresponding author: kumleng_g@yahoo.com Abstract A new three and five step block linear methods based on the Adams family for the direct solution of stiff initial value problems (IVPs) are proposed. The main methods together with the additional methods which constitute the block methods are derived via interpolation and collocation procedures. These methods are of uniform order four and six for the three and five step methods respectively. The stability analysis of the two methods indicates that the methods are A–stable, consistent and zero stable. Numerical results obtained using the proposed new block methods show that they are attractive for the solutions of stiff problems and compete favorably with the well-known Matlab stiff ODE solver ODE23S. Keywords: Linear multistep methods, initial value problems, interpolation and collocation. 1. Introduction In this paper, we shall be concerned with the approximate numerical integration of the stiff initial value problem y ' = f ( x, y ), y ( x0 ) = y0 , x ∈ [ a, b], y ∈ R (1) Numerical analysts have focused most of their works on the development of more efficient and accurate methods for the solution of stiff problems, and as such many methods have been proposed. Notable among these methods is the Backward Differentiation formulae (BDF), because of its wide region of absolute stability. Several researchers such as Kim (2010), Chollom et. al. (2012), Kumleng et.al (2011, 2012) have proposed methods for the numerical solution of (1) which were shown to be very efficient. In this paper, we describe the construction of a new three and five step methods based on the Adams methods similar to the generalized Adams methods of Brugnano and Trigiante (1998). These methods are constructed using the interpolation concept where the continuous interpolants provide the block methods through evaluation at some grid points. This approach produce two self- starting new A – stable block methods which provide the solutions of stiff initial value problems on non-overlapping intervals. 2. The New Method In this section, the new A – stable block methods shall be constructed based on the continuous finite difference approximation approach using the interpolation and collocation criteria described by Lie and Norsett (1981) called multistep collocation (MC) and block multistep methods by Onumanyi et al. (1994). We define based on the interpolation and collocation methods the continuous form of the k- step new method as m yn + v − α v −1 ( x) yn+ v −1 = h∑ β j ( x) f n+ j j =0 where α v−1 and β j ( x) (2) are the continuous coefficients of the method, m is the number of distinct collocation points, h is the step size and v= k −1 2 , k = 3,5, 7... From Onumanyi et. al (1994), we obtain our matrices D and C = D 1 xn+ v −1  1 0 D = 0 1  M 0 1  2 xn+ v −1 L 2 x0 2 x1 L L 2 xm −1 L L m xn + v −1  m  mx0 −1  mx1m −1   M  m −1 mxm −1   −1 as (3) and 1
  • 2. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013  α v −1,1 α v −1,2 C=  M  α v −1,m +1 hβ 0,1 hβ 0,2 hβ 0,m +1 www.iiste.org hβ m−1,1  hβ m −1,2    M  L hβ m−1,m+1  L L (4) respectively. 2.1 Derivation of the Three Step New Block Method In this case, k = 3, v=1, t = 1 and m = 4 and its continuous form expressed in the form of (2) is; y ( x ) = α1 ( x ) yn + h ( β 0 ( x) f n + β1 ( x ) f n +1 + β 2 ( x) f n + 2 + β3 ( x ) f n + 3 ) (5) The matrix D, in (3) becomes  1 xn  0 1 D = 0 1  0 1 0 1  2 xn 3 xn 2 xn 2 xn+1 2 3 xn 2 3 xn +1 2 xn + 2 2 xn +3 2 3 xn + 2 2 3 xn +3 4 xn  3  4 xn  3 4 xn +1   3 4 xn + 2  3  4 xn +3  (6) Using the Maple software, the inverse of the matrix in (6) is obtained and this yields the elements of the matrix C. The elements of the matrix C substituted into (5) yields the continuous formulation of the method as: ( + (− ) λ y( xn + λ ) = yn + −λ − 11λh + 3λh2 − 24 h3 f n + 12 2 3λ 4h 2 3 4 ) + 2λ2 − 8λh3 f n+ 2 + 3h 3 Evaluating (7) at the following points λ = h, constitute the new three step block method. 4 ( λ 2 6h ( 3λ 2 2h ) − 5λ2 + 8λh3 f n +1 6h 3 4 ) λ − 6λh2 + 24 h3 f n +3 3 4 λ = 2h, λ = 3h yields the following discrete methods which ( 9 f n + 19 f n +1 − 5 f n + 2 + f n +3 ) yn + 2 − yn = h ( f n + 4 f n +1 + f n + 2 ) 3 h y n + 3 − y n = 8 ( 3 f n + 9 f n +1 + 9 f n + 2 + 3 f n + 3 ) yn +1 − yn = (7) h 24 (8) The main new discrete scheme for the three step method is yn +1 − yn = h 24 ( 9 f n + 19 f n +1 − 5 f n +2 + f n +3 ) (9) This new method is consistent since its order is 4, it is also zero-stable, above all, it is A – stable as can be seen in figure 1. The new three step discrete methods that constitute the block method (8) have the following orders and error constants as can be seen in Table 1. 2.2 Derivation of the Five –Step New Block Method In this case, k = 5, t = 1 and m = 6 and its continuous form expressed in the form of (2) is; y ( x ) = α1 ( x ) yn +1 + h ( β 0 ( x ) f n + β1 ( x ) f n +1 + β 2 ( x ) f n + 2 + β 3 ( x ) f n + 3 + β 4 ( x ) f n + 4 + β 5 ( x ) f n + 5 ) Similarly, we generate the continuous formulation of the new five step method as; 2 (10)
  • 3. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org ( + (− h + − + − +( h− + − + +(− h + − + − +( h − + − + +(− h + − + − ) 95 λ λ y( xn + λ ) = yn+1 + − 288 h + λ − 137λh + 5λ2 − 17λ3 + 40h4 − 720h5 f n 120 8h 96 h 2 5λ 2 2h 1427 1440 5λ 2 2h 133 240 173 1440 77 λ 3 36 h2 5λ 8h 3 160 71λ 4 96 h3 13λ 5 60 h4 13λ 3 6 h2 49λ 4 48 h3 5 h4 3 λ2 5λ 3 36 h2 10 h 41λ 96 h3 4 7λ4 90 h3 5 6 6 6 6 11λ 120 h4 λ − 144 h5 λ5 λ + 720h5 5 60 h4 ) )f )f )f λ + 144h5 f n+2 λ + 720h5 λ5 ) λ + 144h5 fn +1 7 λ5 60 h4 59λ 4 48h3 61λ 72 h2 2 4 107 λ 3 36 h2 5λ 2 3h 241 720 3 6 6 n +3 n +4 n +5 (11) Evaluating (11) at the following points λ = 0, λ = 2 h, λ = 3h, λ = 4 h, λ = 5h yields the following discrete methods which constitute the new five step block method. h yn + 1 - yn = 1440 ( 475 fn + 1427 fn + 1 - 798 fn + 2 + 482 fn + 3 -173fn + 4 + 27 fn + 5 ) h yn + 2 - yn+1 = 1440 ( -27 fn + 637 fn + 1 +1022 fn + 2 -258 fn + 3 +77 fn + 4 -11 fn + 5 ) h yn + 3 - yn+1 = 90 ( - fn + 34 fn + 1 +114 fn + 2 +34 fn + 3 - fn + 4 ) h yn + 4 - yn+1 = 160 ( -3fn + 69 fn + 1 +174 fn + 2 +174 fn + 3 +69 fn + 4 -3 fn + 5 ) h yn + 5 - yn+1 = 45 (14 fn + 1 +64 fn + 2 +24 fn + 3 +64 fn + 4 +14 fn + 5 ) (12) The main new discrete scheme for the five step method is h yn + 2 - yn +1 = 1440 ( −27 f n + 637 f n + 1 + 1022 f n + 2 − 258 f n + 3 + 77 f n + 4 − 11 f n + 5 ) (13) This new method is consistent since its order is 6, its also zero-stable, above all, it is A – stable as can be seen in figure 1. The new five step discrete methods (12) have the following orders and error constants as shown in Table 1. 3. Analysis of the New Methods In this section, we consider the analysis of the newly constructed methods. Their convergence is determined and their regions of absolute stability are plotted. 3.1 Convergence The convergence of the new block methods is determined using the approach by Fatunla (1991) and Chollom et. al (2007) for linear multistep methods, where the block methods are represented in a single block, r point multistep method of the form k k i =1 i =0 A(0) ym +1 = ∑ A(i ) ym+1 + h∑ B (i ) f m −1 (14) where h is a fixed mesh size within a block, A , B , i = 0,1, 2,...k are i ym , ym −1 and ym +1 i r × r identity matrix while are vectors of numerical estimates. Definition 1: A block method is zero stable provided the roots polynomial ρ ( R ) specified as  k Rij = 1, 2,...k of the first characteristic  ρ ( R) = det ∑ A(i ) R k −i  = 0  i =0 satisfies Rj ≤ 1  (15) , the multiplicity must not exceed two. Fatunla (1994) 3
  • 4. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org The block method (8) expressed in the form of (14) gives  19  1 0 0   yn +1   0 0 1   yn − 2  24  4        0 1 0   yn + 2  =  0 0 1   yn −1  + h  3  0 0 1   y   0 0 1  y   9    n +3    n   8 −5 24   f n +1   0 0    0   fn+2  +  0 0 3   0 0 8   f n +3   1 24 1 3 9 8 9 24 1 3 3 8   f n−2       f n −1       fn   (16) where −5 9 1  19 24 24   0 0 24  1 0 0  0 0 1 24         A(0) =  0 1 0  , A(1) =  0 0 1 , B (0) =  4 1 0  , B (1) =  0 0 1  3 3 3 0 0 1  0 0 1 3  9 9 3      8 8 8 0 0 8  (0) (1) Substituting A and A into (15) gives the characteristic polynomial of the block method (8) as ρ (λ ) = det(λ A(0) − A(1) )   1 0 0   0 0 1       = det λ  0 1 0  −  0 0 1       0 0 1   0 0 1      λ 0 −1  = det  0 λ −1     0 0 λ − 1   = λ 2 (λ − 1) = 0 λ = 1, λ2 = λ3 = 0 . The block method (8) by definition1 is zero stable and by Henrici (1962), the Therefore, 1 block method is convergent since it is also consistent. Similarly, the block method (12) expressed in the form of (14) gives 1   −1  −1   −1  −1  0 0 0 0   yn +1   0    1 0 0 0   yn + 2   0 0 1 0 0   yn + 3  =  0    0 0 1 0   yn + 4   0   0 0 0 1   yn + 5   0    1427 1440  637  1440 h  34 90  69  160  14  45 −798 1440 1022 1440 114 90 174 160 64 45 482 1440 −258 1440 34 90 174 160 24 45 −173 1440 77 1440 −1 90 69 160 64 45  −11  1440  0   −3 160  14   45   27 1440 0 0 0 1   yn − 4    0 0 0 0   yn − 3  0 0 0 0   yn − 2  +   0 0 0 0   yn −1    0 0 0 0   yn   f n +1   0   f n+ 2   0 f n +3  +  0   f n+ 4   0  f n +5   0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0    −1   90  −3 160    0  475 1440 −27 1440 f n− 4    f n −3   f n− 2    f n −1    fn   (17) where 4
  • 5. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 A(0) 1   −1 =  −1   −1  −1  B (0)  1427 1440  637  1440 =  34 90  69  160  14  45 Substituting www.iiste.org 0 0 0 0 0 0   1 0 0 0 0 0 (1) 0 1 0 0, A =  0 0   0 0 1 0 0 0   0 0 0 1 0 0 −798 1440 1022 1440 114 90 174 160 64 45 482 1440 −258 1440 34 90 174 160 24 45 −173 1440 77 1440 −1 90 69 160 64 45 0 0 1  0 0 0 0 0 0  0 0 0  0 0 0  0 0   −11 1440  0 0 (1) 0 , B = 0 0   −3 160  0 0 0 0 14   45  27 1440 0 0 0 0 0 0 0 0 0 0    −1  90  −3 160   0  475 1440 −27 1440 A(0) and A(1) into (15) gives the characteristic polynomial of the block method (12) as ρ (λ ) = det(λ A(0) − A(1) )  1     −1 = det λ  −1     −1   −1   λ  −λ  = det  −λ   −λ  −λ  0 0 0 0 0   1 0 0 0 0 0 1 0 0 − 0   0 0 1 0 0  0 0 0 1 0  0 0 0 λ 0 0 0 λ 0 0 0 λ 0 0 0 0 0 0 1   0 0 0 0  0 0 0 0   0 0 0 0   0 0 0 0   −1 0  0  0 λ  = λ5 − λ4 = 0 λ = 1, λ2 = λ3 = λ4 = λ5 = 0 . The block method (12) by definition1 is zero stable and by Henrici Therefore, 1 (1962), it is convergent since it is also consistent. 3.2 Regions of Absolute Stability of the Methods The absolute stability regions of the newly constructed block methods (8) and (12) are plotted using Chollom (2005) by reformulating the methods as general linear methods. The regions of absolute stability of the methods of the main discrete schemes and the new block methods are as shown in the figures 1 and 2 below. These absolute stability regions are all A –stable since they consist of the set of points in the complex plane outside the enclosed figures. 4 Numerical Examples We report here a few numerical examples on some stiff problems taken from the literature. For comparisons, we also report the performance of the new block methods and the well-known Matlab stiff ODE solver ODE23S on the same problems. 5
  • 6. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org Problem 1: We consider the Robertson’s problem y '1 = −0.04 y1 + 10000 y2 y3 y1 (0) = 1 2 y '2 = 0.04 y1 − 10000 y2 y3 − 30000000 y2 y2 (0) = 0 2 y '3 = 30000000 y2 y3 (0) = 0 0 ≤ x ≤ 70, h = 0.1 Problem 2: We consider the Van der Pol’s equation y '1 = y2 y1 (0) = 2 y '2 = - y1 + µ y2 (1- y12 ) , y2 (0) = 0 0 ≤ x ≤ 70, h = 0.01, µ = 10, 5. Conclusion The new block methods were applied to two well-known stiff problems from the literature, the numerical results suggest that the proposed new block methods (8) and (12) are suitable for solving stiff problems and perform competitively with the well-known ODE23s. This success is achieved because of the good stability properties of the proposed new block methods. References Brugnano, L. and D, Trigiante. (1998). Solving Differential Problems by Multistep Initial and Boundary Value Methods. Gordon and Breach Science Publishers, Amsterdam. Chollom, J.P., (2005). The construction of Block Hybrid Adam Moulton Methods with link to Two Step RungeKutta Methods. (Ph.D) Thesis University of Jos. Chollom, J.P., Ndam, J.N. and Kumleng, G.M. (2007). On Some Properties of the Block Linear Multistep Methods. Science World Journal, 2(3), 11 – 17. Chollom, J.P., Olatunbasun,I, O. and Omagu, S. (2012). A Class of A-Stable Block Explicit Methods for the Solutions of ODEs Research Journal of Mathematics and Statistics, 4(2): 52-56. Fatunla, S.O. (1991). Parallel Method for Second Order Differential Equations. Proceedings of the National Conference on Computational Mathematics, University of Ibadan Press, pp:87-99. Fatunla, S.O. (1994). Block Method for Second Order Differential Equations. International Journal of Computer Mathematics. 41,55 – 63. Henrici, P. (1962). Discrete Variable Methods in Ordinary Differential Equations. John Wiley, New York, p407 Kim, P. (2010). An Explicit Type Stable Method for Solving Stiff Initial Value Problems. Presentation at a mini workshop at Knu. Republic of Korea, Retrieved from: http://webbuild.knu.ac.kr/~skim/conf_math2/kimps.pdf. Kumleng,G.M. and Skwame,Y. (2011) A New A-Stable Method for the Solution of Stiff Initial Value Problems. International Journal of Numerical Maths.6 (2), 360 -373. Kumelng,G.M., Sirisena, U.W. and Chollom, J.P.(2012) Construction of a Class of Block Hybrid Implicit Multistep Methods for the Solution of Stiff Ordinary Differential Equations. Nigerian Journal of Pure and Applied Sciences 5. Lie, I. and Norsett, S.P. (1989). Super convergence for Multistep Collocation. Mathematics of Computation. 52(185), 65 – 79. Onumanyi, P., Awoyemi, D.O., Jator, S.N. and Sirisena, U.W. (1994).New Linear Multistep Methods with Continuous Coefficients for First Order Initial Value Problems. Journal of the Nigerian Mathematical Society .13, 37 – 51. 6
  • 7. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org 2.5 2 1.5 1 I m (z ) 0.5 0 k=3 k=5 -0.5 -1 -1.5 -2 -2.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Re(z) Figure2. Absolute Stability Regions of the new Three and Five step discrete methods 2.5 2 1.5 1 I m (z ) 0.5 0 k=5 k=3 -0.5 -1 -1.5 -2 -2.5 -0.5 0 0.5 1 Re(z) 1.5 2 2.5 Figure1. Absolute Stability Regions of the Three and Five step New block methods The absolute stability regions consists of the set of points in the complex plane outside the enclosed figures. Therefore, both the discrete and block methods are all A - stable since the left –half complex plane is contained in S., where S = { z ∈ C : R ( z ) ≤ 1} . 7
  • 8. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org 1 y1 Num y2 Num y3 Num y1 ODE23s y2 ODE23s y3 ODE23s 0.9 0.8 0.7 S o lu t io n (y ) 0.6 0.5 0.4 0.3 0.2 0.1 0 0 10 20 30 40 50 60 70 Time(t) Figure 1. Solution to Problem 1 using the Three Step Block Method and ODE23S y1 Num y2 Num y3 Num y1 ODE23s y2 ODE23s y3 ODE23s 1 0.9 0.8 S o lu tio n (y ) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 10 20 30 40 50 60 70 Time(t) Figure 2. Solution to Problem 1 using the Five Step Block Method and ODE23S 15 y1 Num y2 Num y1 ODE23s y2 ODE23s 10 S o lu t io n ( y ) 5 0 -5 -10 -15 0 10 20 30 40 50 60 70 Time(t) Figure 3. Solution to Problem 2 using the Three Step Block Method and ODE23S 8
  • 9. Mathematical Theory and Modeling ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.3, No.11, 2013 www.iiste.org 15 y1 Num y2 Num y1 ODE23s y2 ODE23s 10 S o lu t io n ( y ) 5 0 -5 -10 -15 0 10 20 30 40 50 60 70 Time(t) Figure 4. Solution to Problem 2 using the Five Step Block Method and ODE23S Table 1: Order and Error Constants of the new three step method Method y n +1 Order 4 Error constant − 4 y n+2 1 − 90 3 − 80 4 y n+3 19 720 Table 2: Order and Error Constant of the new five step method Method y n +1 y n+2 y n+3 Order 6 Error constant 863 − 60480 271 60480 6 1 756 6 y n+4 y n+5 13 2240 6 6 8 − 945 9
  • 10. This academic article was published by The International Institute for Science, Technology and Education (IISTE). The IISTE is a pioneer in the Open Access Publishing service based in the U.S. and Europe. The aim of the institute is Accelerating Global Knowledge Sharing. More information about the publisher can be found in the IISTE’s homepage: http://www.iiste.org CALL FOR JOURNAL PAPERS The IISTE is currently hosting more than 30 peer-reviewed academic journals and collaborating with academic institutions around the world. There’s no deadline for submission. Prospective authors of IISTE journals can find the submission instruction on the following page: http://www.iiste.org/journals/ The IISTE editorial team promises to the review and publish all the qualified submissions in a fast manner. All the journals articles are available online to the readers all over the world without financial, legal, or technical barriers other than those inseparable from gaining access to the internet itself. Printed version of the journals is also available upon request of readers and authors. MORE RESOURCES Book publication information: http://www.iiste.org/book/ Recent conferences: http://www.iiste.org/conference/ IISTE Knowledge Sharing Partners EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open Archives Harvester, Bielefeld Academic Search Engine, Elektronische Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial Library , NewJour, Google Scholar