Scribd
Upload
53 views8 pages

Sec5 3 Taylor

This document discusses high-order Taylor methods and Runge-Kutta methods for solving initial value problems (IVPs) numerically. It defines the local truncation error of a difference method and shows that Euler's method has an error of O(h). Higher-order Taylor methods are derived by expanding the solution in a Taylor polynomial and discarding higher-order terms. Runge-Kutta methods aim to achieve high-order accuracy without needing derivatives by applying Taylor's theorem in two variables. A second-order Runge-Kutta method called the midpoint method is derived as an example.

Uploaded by

Trương Đoàn Tuấn Kiệt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
53 views8 pages

Sec5 3 Taylor

This document discusses high-order Taylor methods and Runge-Kutta methods for solving initial value problems (IVPs) numerically. It defines the local truncation error of a difference method and shows that Euler's method has an error of O(h). Higher-order Taylor methods are derived by expanding the solution in a Taylor polynomial and discarding higher-order terms. Runge-Kutta methods aim to achieve high-order accuracy without needing derivatives by applying Taylor's theorem in two variables. A second-order Runge-Kutta method called the midpoint method is derived as an example.

Uploaded by

Trương Đoàn Tuấn Kiệt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

5.

3 High-Order Taylor Methods


Consider the IVP

( )

Definition: The difference method


(

has Local Truncation Error

Note:

( ) and

( )
(

))

).

Geometric view of local truncation error

( )

Example.
( )
Solution:

Analyze

with
( )

( )

( )

the

local
. Assume |

))

truncation
error
of
Eulers
( )|
with
constant.

( )

( )

method

for

solving

).
( ).

( )|
Thus |
.
So the local truncation error in Eulers method is ( ).

Consider the IVP


Compute
First, by IVP:

( )

(
.

( )

( )

( )

( )

( ))
( ( ))

(
( )

Second, by chain rule:


( )
(
)

( ))

( ))

( ))

( ))

( )

( ))

( ))

( ))

Derivation of higher-order Taylor methods


Consider the IVP
(

( )
2

Expand ( ) in the nth Taylor polynomial about , evaluated at


(
( )
(

for some
Denote

( )

( )

( )

( ))

( )

( )
(

( ))

( )

( ))

)
(

( )

( )

( ))

). Delete remainder term to obtain the nth Taylor method of order n.


)

Taylor method of order n


( )

Remark: Eulers method is Taylor method of order one.


Example 1. Use Taylor method of orders (a) two and (b) four with N = 10 to the IVP
( )
Solution:
. So
(a)

( ))

)
(
)
So ( ) (
nd
The 2 order Taylor method is

( (
Now compute approximations at each time step:

)
3

(
(

(
(

(b)

( )

( ))

( )(

( ))
( )

( )
( )

)
)

|
|

)
)(

|
|

( ))
( ))

(
)(

( )

)
)

The 4th order Taylor method is


((

)(

)(

.
Now compute approximate solutions at each time step:
((

)(

)( )

(
|

((

)(

|
)(

(
|

))

|
4

Finding approximations at time other than


( )
Example. (Table 5.4 on Page 259). Assume the IVP
is solved by the 4th order
)
Taylors method with time step size
.
(
),
(
). Find (
Solution:
Method 1: use linear Lagrange interpolation.
(
)
Method 2: use Hermite polynomial interpolation (more accurate than the result by linear Lagrange interpolation).
First use
to approximate ( ) and ( )
( )
( ) ( )
( )
( )
( ) ( )
( )
Then use Theorem 3.9 to construct Hermite polynomial ( ).
(
)
(
)

Error analysis
Theorem 5.12 If Taylor method of order n is used to approximate the solution to the IVP
( )
( )
with step size h and if
, then the local truncation error is ( ).
Remark: (

( )

( ))

( )

( )
(

Thus |

( )

( )

( )|

(
(

( )) is bounded by |
)

( ))

( )|

(
)

( ))
( )

( )
(

( ))

( ))

So the local truncation error in Eulers method is (

).

5.4 Runge-Kutta Methods


Motivation: Obtain high-order accuracy of Taylors method without knowledge of derivatives of (

).

Theorem 5.13(Taylors Theorem in Two Variables) Suppose ( ) and partial derivative up to order
( )|
, let (
)
. For ( )
, there is
and
with
( )
( )
( )
Here
(

)
[

( )(

[(

)
)

)
(

)
)(

)]

(
(

)(

continuous on

)]

)]

( ))

) is the nth Taylor polynomial in two variables.

Derivation of Runge-Kutta method of order two


1. Determine , , such that
(
Notice

We have
2. Expand

)
( )

)
(

( ))

)
(

( ))
(

( ))

(
(

( ))
(

( )
( ))

( )

( ))

( ))

(
(

( ))

( )

) in 1st degree Taylor polynomial:


6

( )

3. Match coefficients of equation (1) and (2) gives


(

( ))

with unique solution


(

( ))

4. This gives
( )

with

( )))

( )))

Local truncation error is (


Two stage formula:

( )))

Midpoint Method (one of Runge-Kutta methods of order two)


Consider the IVP
( )
with step size
(

( )

))

)
(
(

)
)

Example 2. Use the Midpoint method with

and

to solve the IVP


( )

You might also like