Numerical solutions to ODE

There exists large number of ordinary differential equations, whose solution cannot be obtained
by the known analytical methods. In such cases, we use numerical methods to get an
approximate solution of a given differential equation with given initial condition.

Consider an ordinary differential equation of first order and first degree of the form
dy
 f  x, y  with the initial condition y  x0   y0 which is called initial value problem.
dx
 Euler’s Method
dy
Consider the first order differential equation  f  x, y 
dx
With the initial condition y  x0   y0
The approximate value of y at x  x1 is denoted by y1
y1  y0  hf  x0 , y0 
The approximate value of y at x  x2 is denoted by y2
y2  y1  hf  x1 , y1 
In general, we obtain a recursive relation as
yn 1  yn  hf  xn , yn  n = 0, 1, 2, ………
This is known as Euler’s algorithm
 dy
Using Euler's method, solve for y at x  0.4, from  e , y 0  0
x2

dx
taking step size h  0.1
dy
Sol: Given that  e if y  0   0 and h  0.1
x2

dx
dy
 f  x, y   e with initial condition y  0   0
x2

dx
The Euler's method is yn 1  yn  hf  xn , yn 
 i  y1  y0  hf  x0 , y0   0  0.1 f  0, 0   0  0.11  0.1
y1  y  0.1  0.1
 ii  y2  y1  hf  x1 , y1   0.1  0.1 f  0.1, 0.1  0.1  0.11.0100   0.2
y2  y  0.2   0.2
 iii  y3  y2  hf  x2 , y2   0.2  0.1 f  0.2, 0.2   0.2  0.11.0408   0.3
y3  y  0.3  0.3
 iv  y4  y3  hf  x3 , y3   0.3  0.1 f  0.3, 0.3  0.3  0.11.0941  0.41
y4  y  0.4   0.41
 dy y  x
Problem: Given  with initial condition y  0   1, find y  0.1 by Euler's
dx y  x
method, taking step size h  0.02
yx
Sol: Given f  x, y  
yx
First we divide the x values into sub intervals using h
x0  0, x1  0.02, x2  0.04, x3  0.06, x4  0.08, x5  0.1
y0  0, y1 , y2 , y3 , y4 , y5  0.1
y1  y0  hf  x0 , y0   1  0.02 f  0,1  1  0.02 1  1.02
y1  y  0.02   1.02
y2  y1  hf  x1 , y1   1.02  0.02 f  0.02,1.02   1.02  0.02  0.9615   1.039
y2  y  0.04   1.039
y3  y2  hf  x2 , y2   1.039  0.02 f  0.04,1.039   1.039  0.02  0.9258   1.057
y3  y  0.06   1.057
y4  y3  hf  x3 , y3   1.057  0.02 f  0.06,1.057   1.057  0.02  0.8925   1.075
y4  y  0.08   1.075
y5  y4  hf  x4 , y4   1.075  0.02 f  0.08,1.075   1.075  0.02  0.8615   1.092
y5  y  0.1  1.092
 Modified Euler’s method
dy
Consider the first order differential equation  f  x, y 
dx
with the initial condition y  x0   y0
The approximate value of y at x  x1 is denoted by y1
y1(0)  y0  hf ( x0 , y0 )
h
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1( 0) ) ,
(1)

2
h
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1(1) ) , ….,
( 2)

2
( K 1) h
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1( K ) )
2
To get y2 we use the above procedure again.
Using modified Euler’s method find y  0.1 and y  0.2 
dy
given that  x2  y, y  0   1 , taking step size h = 0.1
dx
dy
Sol. Here  x2  y, y  0   1 and h = 0.1
dx
To find y1 i .e y(0.1)
Using Modified Euler’s formula
y1(0)  y0  hf ( x0 , y0 )  1   0.1 1  0.9
The First Approximation to y1
h 0.1
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1 )  1 
(1) ( 0)
(1  f (0.1,0.89))  0.995
2 2
Second Approximation to y1
h 0.1
y1( 2)  y0  ( f ( x0 , y0 )  f ( x1 , y1(1) ) = 1+ ((1  f (0.1,0.995))  0.90075
2 2
Third Approximation to y1
h 0.1
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1 ) = 1+
( 3) ( 2)
((1  f (0.1,0.90075 ))  0.90546
2 2
Fourth Approximation to y1
h 0.1
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1 )  1 
( 4) ( 3)
((1  f (0.1,0.90546 ))  0.90523
2 2
Fifth Approximation to y1
h 0.1
y1  y0  ( f ( x0 , y0 )  f ( x1 , y1 )  1 
(5) (4)
((1  f (0.1, 0.90523))
2 2
0.1
 1 (1  0.89523)  0.90523
2
Since y1( 4) , y1(5) are same, therefore y  0.1 = 0.90523
To find y2 i.e y(0.2)
Using Modified Euler’s formula
y2(0)  y1  h( f ( x1 , y1 ) = 0.90523 + 0.1 f(0.1, 0.90523) = 0.90523+0.1(-0.8952)=0.8157
The First Approximation to y2
h 0.1
y2(1)  y1  ( f ( x1 , y1 )  f ( x2 , y2(0) )  0.90523  (0.8952  f (0.2, 0.8157))
2 2
0.1
 0.90523  (0.8952  (0.04  0.8157))  0.90523  0.0835  0.8217
2
The Second Approximation to y2
h 0.1
y2(2)  y1  ( f ( x1 , y1 )  f ( x2 , y2(1) )  0.90523  (0.8952  f (0.2, 0.8217))
2 2
 0.90523  0.08385  0.8214
The Third Approximation to y2
h 0.1
y2(3)  y1  ( f ( x1 , y1 )  f ( x2 , y2(2) )  0.90523  ( 0.8952  f (0.2, 0.8214))
2 2
 0.90532  0.08383  0.8214
Since y2( 2) , y2(3) are same, therefore y  0.2  = 0.8214
Runge - Kutta method of fourth order
dy
Consider the first order differential equation  f  x, y 
dx
with the initial condition y  x0   y0
The Runge - Kutta fourth order method is

k1  hf  x0 , y0 
 h k 
k2  hf  x0  , y0  1 
 2 2
 h k 
k3  hf  x0  , y0  2 
 2 2
k4  hf  x0  h, y0  k3 
1
Now k   k1  2k2  2k3  k4 
6
Then we can calculate y1  y0  k
Using Runge -Kutta method of order four, find y  0.2 
dy
from  f  x, y   y  x , y  0   2, taking step size h  0.1
dx
dy
Sol: Given  f  x, y   y  x , x0  0, y0  2, h  0.1
dx
k1  hf  x0 , y0   0.1 f  0, 2   0.1  2  0   0.2
 h k1   0.1 0.2 
k2  hf  x0  , y0    0.1 f  0  ,2 
 2 2  2 2 
 0.1 f  0.05, 2.1  0.1 2.1  0.05   0.205
  h k2   0.1 0.205 
k3  hf  x0  , y0    0.1 f  0  ,2 
 2 2  2 2 
 0.1 f  0.05, 2.1025   0.1 2.1025  0.05   0.20525
k4  hf  x0  h, y0  k3   0.1 f  0  0.1, 2  0.20525 
 0.1 f  0.1, 2.20525   0.1 2.20525  0.1  0.210525
1
then k   k1  2k2  2k3  k4 
6
1
  0.2  2  0.205  2  0.20525  0.210525   0.11607
6
and y1  y0  k  2  0.11607  2.11607
 ii  Now x1  0.1, y1  2.11607, h  0.1
k1  hf  x1 , y1   0.1 f  0.1, 2.11607 
 0.1  2.11607  0.1  0.201607
 h k1   0.2 0.201607 
k2  hf  x1  , y1    0.1 f  0.1  , 2.11607  
 2 2   2 2 
 0.1 f  0.2, 2.168735   0.20169
 h k2   0.2 0.20169 
k3  hf  x1  , y1    0.1 f  0.1  , 2.11607  
 2 2   2 2 
 0.1 f  0.2, 2.21691  0.20169
k4  hf  x1  h, y1  k3   0.1 f  0.1  0.1, 2.11607  0.20169 
 0.1 f  0.2, 2.21692 
 0.20169
1
then k   k1  2k2  2k3  k4 
6
1
  0.201607  2  0.20169  2  0.20169  0.20169 
6
 0.20168
and y2  y1  k  2.11607  0.20168  2.31775