MTH2212 – Computational Methods and

Statistics

Numerical Solutions of Ordinary

Differential Equations

Lecture 12:
Initial Value Problem: Runge-Kutta Methods
 Objectives

 Introduction
 Runge-Kutta Methods
 Second-order Runge-Kutta
Heun Method
Midpoint Method
Ralston’s Method
 Fourth-order Runge-Kutta
 System of ODE

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 2

 Introduction

 A fundamental source of error

in Euler’s method is that the
derivative at the beginning of
the interval is assumed to
apply across the entire
interval.
 Improving the estimate of the
slope will improve the
solution.

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 3

 Runge-Kutta Methods

 Runge-Kutta (RK) methods achieve the accuracy of a Taylor

series without requiring the calculation of higher
derivatives.
 For a differential equation of the form:
dy
 f ( x, y )
dx
The estimate of the solution is given by
yi 1  yi  ( xi , yi , h) h

where ( xi , yi , h) is the increment function which is the

slope over the interval
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 4
 Runge-Kutta Methods

 The increment function can be written in general form as

  a1k1  a2 k 2  a3 k 3  ...  an k n
where a1, a2, a3,…, an are constants and k1, k2, k3,…, kn are
k1  f ( xi , yi )
k 2  f ( xi  p1h, yi  q11 k1 h)
k 3  f ( xi  p 2 h, yi  q21k1 h  q22 k 2 h)
.
.
.
k n  f ( xi  pn 1h, yi  q n1,1k1 h  qn1, 2 k 2 h  ...  qn 1,n 1k n1h)
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 5
 Runge-Kutta Methods

 Where p1, p2,…,pn-1 and q1, q2,…,qn-1 are all constants

 n is order number of the RK method
 k’s are recurrence functions. Because each k is a functional
evaluation, this recurrence makes RK methods efficient for
computer calculations.
 Various types of RK methods can be devised by employing
different number of terms in the increment function as
specified by n.
 Once n is chosen, values of a’s, p’s, and q’s are evaluated by
setting general equation equal to terms in a Taylor series
expansion.
 First-order RK method (n = 1) is Euler’s method.

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 6

 Second-order Runge-Kutta

 The second-order version of Runge-Kutta methods is

yi 1  yi  (a1k1  a 2 k 2 ) h (i)
where
k1  f ( xi , yi )
k 2  f ( xi  p1 h, yi  q11 k1h)
 Evaluate values for a1, a2, p1 and q11 by setting equation (i) equal
to a Taylor series expansion to the second-order term (see Box
25.1 p.703 for details).
 Three equations are derived to solve four unknown constants
a1  a2  1
a2 p1  1 / 2
a2 q11  1 / 2
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 7
 Second-order Runge-Kutta

 Since there are three equations with four unknowns, we

must assume a value of one of the unknowns to determine
the other three.
 Suppose that we specify a value for a2, then
a1  1  a2
1
p1  q11 
2a 2
 There are an infinite number of second-order RK methods

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 8

 Second-order Runge-Kutta

 Because we can choose an infinite number of values for a2,

there are an infinite number of second-order RK methods.
 Every version would yield exactly the same results if the
solution to ODE were quadratic, linear, or a constant.
 However, they yield different results if the solution is more
complicated (typically the case).
 Three of the most commonly used RK second-order
methods are:
 Heun’s Method with a Single Corrector (a2=1/2)

 The Midpoint Method (a2=1)

 Raltson’s Method (a =2/3)

2

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 9

 Heun Method with a Single Corrector

 a2 is set to a2=1/2
a1  1  a2  1 / 2
p1  q11  1 / 2a 2
Substituting into equation (i)
1 1
yi 1  yi  ( k1  k 2 ) h
Where 2 2
k1  f ( xi , yi )
k 2  f ( xi  h, yi  k1h)
 k1 is the slope at the beginning of the interval and k2 is the
slope at the end of the interval.

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 10

 The Midpoint Method

 a2 is set to a2=1
a1  0
1
p1  q11 
2
 Substituting into equation (i)
yi 1  yi  k 2 h

Where
k1  f ( xi , yi )
1 1
k 2  f ( xi  h, yi  k1h)
2 2
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 11
 Ralston’s Method

 a2 is set to a2=2/3
a1  1 / 3
p1  q11  3 / 4

Substituting into equation (i)

1 2
yi 1  yi  ( k1  k 2 ) h
3 3
Where
k1  f ( xi , yi )
3 3
k 2  f ( xi  h, yi  k1 h)
4 4
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 12
 Second-order Runge-Kutta

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 13

 Fourth-order Runge-Kutta

 This is the most popular of the RK methods

1
yi 1  yi  (k1  2k 2  2k 3  k 4 ) h
6
Where
k1  f ( xi , yi )
1 1
k 2  f ( xi  h, yi  k1 h)
2 2
1 1
k 3  f ( xi  h, yi  k 2 h)
2 2
k 4  f ( xi  h, yi  k 3 h)

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 14

 Example 1

Integrate using fourth-order Runge-Kutta method:

dy
 4e0.8 x  0.5 y
dx
Using h = 0.5 with y(0) = 2 from x = 0 to 0.5.

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 15

 Example 1 - Solution

 Compute the slope at the beginning of the interval:

k1  f ( x i , yi )  f (0,2)  4e 0.8( 0 )  0.5(2)  3

 Use this slope to compute the value of y and slope at midpoint

y (0.25)  2  3(0.25)  2.75
k 2  f (0.25,2.75)  4e 0.8( 0.25 )  0.5(2.75)  3.510611

 Use this slope to calculate another value of y and slope at the

midpoint
y (0.25)  2  3.510611(0.25)  2.877653
k3  f (0.25,2.877653)  4e 0.8( 0.25)  0.5(2.877653)  3.446785

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 16

 Example 1 - Solution

 Use the this slope to compute the slope at the end of the
interval:
y (0.5)  2  3.071785(0.5)  3.723392

k 4  f (0.5,3.723392)  4e 0.8( 0.5)  0.5(3.723392)  4.105603

 Evaluate the final prediction at the end of the interval

1
yi 1  yi  (k1  2k 2  2k3  k 4 ) h
6
1
y (0.5)  2  (3  2  3.510611  2  3.446785  4.105603) 0.5
6
 3.751669

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 17

 System of Equations

 Many practical problems in engineering require the solution of a

system of simultaneous ODE rather than a single equation
 dy1
 dx  f1 ( x, y1 , y2 ,..., yn )

 dy2  f ( x, y , y ,..., y )
 dx 2 1 2 n


.
.

.
 dy
 n  f n ( x, y1 , y2 ,..., yn )
 dx
 Solution requires that n initial conditions be known at the starting
value x.

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 18

 Procedure for Solution Using RK Methods

1. Compute the slopes for all variables at the beginning of the

interval (set of k1’s).

2. Use these slopes to make predictions of the dependent

variable at the midpoint of the interval (y1, y2,…). Use the
midpoint values to compute a first set of slopes at the
midpoint of the interval (set of k2’s)

3. The new set of slopes k2 are used to make another set of

midpoint predictions (y1, y2,…). Use the latest set of
midpoint predictions (y1, y2,…) to compute the second set
of midpoint slopes (set of k3’s)

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 19

 Procedure for Solution Using RK Methods

4. Use the values of k3’s to make predictions at the end of the

interval (y1, y2,…). Use the predictions at the end of the
interval to compute the endpoint slopes (set of k4’s)

5. Combine the k’s into a set of increment functions and

brought back to the beginning to make the final prediction
of (y1, y2,…).

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 20

 Example 2

Solve the following set of differential equations using the 4th

order RK method, assuming that at x = 0, y1 = 4 and y2 = 6.
Integrate to x = 2 with a step size of 0.5

 dy1  dy1
 dx  0.5 y  dx  0.5 y  f1 ( x, y1 , y 2 )
 
 dy 2  4  0.3 y  0.1 y  dy 2  4  0.3 y  0.1 y  f ( x, y , y )
 dx 2 1
 dx 2 1 2 1 2

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 21

 Example 2 - Solution

1. Compute the slopes for all variables at the beginning of the

interval (set of k1’s).
k1,1  f1 (0,4,6)  0.5(4)  2
k1, 2  f 2 (0,4,6)  4  0.3(6)  0.1(4)  1.8
2. Use set of k1’s to make predictions of (y1, y2,…) at the midpoint
of the interval. Use these (y1, y2,…) to compute a first set of
slopes at the midpoint of the interval (set of k2’s)
h
y1 (0.25)  y1 (0)  k1,1  4  (2)(0.25)  3.5
2
h
y2 (0.25)  y2 (0)  k1, 2  6  (1.8)(0.25)  6.45
2
k 2,1  f1 (0.25,3.5,6.45)  0.5(3.5)  1.75
k 2, 2  f 2 (0.25,3.5,6.45)  4  0.3(6.45)  0.1(3.5)  1.715

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 22

 Example 2 - Solution

3. Use the set of k2’s to make another set of midpoint

predictions (y1, y2,…). Use the latest set of (y1, y2,…) to
compute the second set of midpoint slopes (set of k3’s)
h 0.5
y1 (0.25)  y1 (0)  k2,1  4  (1.75)( )  3.5625
2 2
h 0 .5
y2 (0.25)  y2 (0)  k2, 2  6  (1.715)( )  6.42875
2 2
k3,1  f1 (0.25,3.5625,6.42875)  0.5(3.5625)  1.78125
k3, 2  f 2 (0.25,3.5625,6.42875)  4  0.3(6.42875)  0.1(3.5625)  1.715125

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 23

 Example 2 - Solution

4. Use the values of k3’s to make predictions at the end of the

interval (y1, y2,…). Use the predictions at the end of the interval
to compute the endpoint slopes (set of k4’s)
y1 (0.5)  y1 (0)  k3,1h  4  (1.78125)(0.5)  3.109375
y2 (0.5)  y2 (0)  k3, 2 h  6  (1.715125)(0.5)  6.857563
k 4,1  f1 (0.5,3.109375,6.857563)  0.5(3.109375)  1.554688
k 4, 2  f 2 (0.5,3.109375,6.857563)  4  0.3(6.857563)  0.1(3.109375)  1.631794

5. Combine the k’s into a set of increment functions and brought

back to the beginning to make the final prediction of (y1, y2,…).
1
y1 (0.5)  4  [( 2)  2(1.75)  2(1.78125)  (1.554688)] 0.5  3.115234
6
1
y2 (0.5)  6  [(1.8)  2(1.715)  2(1.715125)  (1.631794)] 0.5  6.857670
6
Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 24
 Example 2 - Solution

 Repeating the procedure for the remaining steps gives the

following results

x y1 y2
0.0 4 6
0.5 3.115234 6.857670
1.0 2.426171 7.632106
1.5 1.889523 8.326886
2.0 1.471577 8.946865

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 25

 Assignment # 5

 Computational Methods
 22.13, 25.4, 25.8, 25.9, 28.33
 Statistics
 Check with Dr Faiz

Dr. M. Hrairi MTH2212 - Computational Methods and Statistics 26