Scribd
Upload
108 views

Errors True Error True Percent Relative Error: Appendix

The document discusses various numerical analysis techniques including: 1) Methods for solving equations such as bisection, Newton-Raphson, and secant methods. 2) Techniques for solving systems of linear equations like Gaussian elimination, Gauss-Seidel, and Jacobi methods. 3) Polynomial interpolation approaches including Newton and Lagrange interpolation. 4) Numerical integration techniques like the trapezoidal rule, Simpson's rule, and composite Simpson's rule. 5) Methods for solving ordinary differential equations including Euler's method and 4th order Runge-Kutta.

Uploaded by

Sin Chien
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
108 views

Errors True Error True Percent Relative Error: Appendix

The document discusses various numerical analysis techniques including: 1) Methods for solving equations such as bisection, Newton-Raphson, and secant methods. 2) Techniques for solving systems of linear equations like Gaussian elimination, Gauss-Seidel, and Jacobi methods. 3) Polynomial interpolation approaches including Newton and Lagrange interpolation. 4) Numerical integration techniques like the trapezoidal rule, Simpson's rule, and composite Simpson's rule. 5) Methods for solving ordinary differential equations including Euler's method and 4th order Runge-Kutta.

Uploaded by

Sin Chien
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/ 5

BEE/BEPIBECIBMMIBMFIBMA/BMIIBMB/BAA/

BAE/141511/BUM2313

CONFIDENTIAL

APPENDIX
Errors
True percent relative error

True Error
= true value - approximation value

true value - approximation value

true value
Approximate percent relative error
= present approximation - previous approximation
present approximation

f(x)(x, xv)

(x +x)

Xr

f(x,)f(x)

Newton-Raphson method

Secant method
x . '

x.

Stopping criterion
Terminate computation when Jt a I <e

Roots of Equations
False-position method

Bisection method
Xr =

X100

xlOO

f(x,)(x,_1 x,)

- f(x1)
f'(x)

' f(x1_1 ) - f(x)

Linear Algebraic Equations and Matrices


System of linear algebraic equations
[A] {X} = {B}.

Decomposition [A] = [L] [U] with [U] and [U] can be obtained as follo ws :

Using Doolittle decomposition


1
[U] = 121

0
1

131

132

0
0 ;[U]
1

U11

0
0

U12

U13

U22

U23

U33

Using Cholesky method

________

a,1

U,, =
11

[A] = [U]T [U]

[U] =

uki

12

u22

u23 ,

i-i

ay
k=i

U/i

for =i+ 1,...,n

CONFIDENTIAL

BEE/BEP/BEC/BMM/BMFIBMA/BMI/BMBIBAA/
BAE/141511/11UM2313

Using Crout's method


[ a11

a12

a13

a141


111

a21

a22

a23

a24

I a31

a32

a33

a34

[ a41

a42

a43

a44j

u12

u13

u141

0 I 0

u23

u24

133

Ii 0

u34

143

144 ][0

122

131

132

[ 141

142

Gauss seidel method

Jacobi method

L_

X+1)

1j

(j=n

a1, [

Ii xII'
j
II

i=1 2,...,n.

a11

a11

(k)

a11

where

k = 1,2,...

Nonlinear system: Newton-Raphson method

Power method
v"' =AvW

072,

fi i-- - f -

m k+l

k = 0,1,2

x111 = x11

il
f2,i

t3Ai

--

x2.i+1 = x2,1

r,
[J]

Newton interpolation polynomial


f. W = f(x0)+f[x0,x1](x-x0)

II
1i

ax,

ax2 I

af ,,

af211

Curve Fitting
Lagrange interpolation polynomial
11

XX
f (x)=L,(x)f(x1) where L1(x)=JJ_ '
i=0
n-

where
=
xn - xo

L1 -i--

order of interpolation

j=O
j#i

X1 -X

CONFIDENTIAL

BEE/BEP/BECIBMM/BMFIBMAJBMIIBMB/BAA/
BAE/141511JBUM2313

Inverse Newton interpolation polynomial


1(f)

=x

+L, (ff0 )+b, (ff0)(fJ)+b, (fJ)(fj)(ff2)

n - order of interpolation
Inverse Lagrange interpolation polynomial
where L(f)=fJ
J^I

n - order of interpolation
Linear Splines
s,(x)= AX) +'(xx,)

Quadratic Splines
s.(x)=a,+b,(xx)+c1(xx,)2
x,
For i=1,2,...,n - 1, find
f+b1 h+c1h,2 =J^
h, =x, 1 x;

x:!^x11

b.+2c,h, =k+1;
Also given,
Cl = 0
a, =
Numerical Integration
Simpson's 113 Id rule

Trapezoidal rule
I[f(xO)+21f(x)+f(x)]

f(x,)+2I f(xi)+f(Xn)

I[f(xo)+4
,135

where
/_
-

where

X. _X0
n

h =

x " xo

and n must even segment

Simpson's 3/8 rule


= X3 x0
[f(x0)+3f(x1)+3f(x2)+f(x3)], where h
Ordinary Differential Equations (IVP)

Euler's method
.+' =y,+hf(x,,y)

= x, + h

CONFIDENTIAL

BEE/BEP/BEC/BMM/BMFIBMA/BMIIBMBIBAAI
BAE/141511/BUM2313

2 nd order Runge-Kutta: Midpoint method

2 order Runge-Kutta: Henn method


1

y1+1 =y1+k2h
= x, + h

v11

=+h

where

where

kj =f(x1,y)

In =f(x1,y)

k2 = f (x, +-1-h,y1 +--Inh)

in2 =f(x1+h,y1+k1h)

order Runge-Kutta: Ralston's method


)i+l = Yi +(k +2k2)h
= x, +h
where
1n =f(x1,y1)
k2 =f(x1+.-h,y1+-k1h)

Fourth order Runge-Kutta method


Y,+1 = y,

+(kj +2k2 +2k2 +k4)

x11 =x, +h

where
1n =f(x1,y1)
k2 = f(x1 +h,y1 +!k1h)
k3 =f (x, +!h,y, +1 k2h
in4 =f(x1+h,y+k2h)

BEE/BEPIBEC/BMM/BMFIBMAIBMIIBMB/BAA/
BAE/14151I1BUM2313

CONFIDENTIAL

Ordinary Differential Equations (BVP)


Shooting method
Extrapolate estimate for initial slope
z(0) = GI + G2 - GI (D - RI)
R2R1

where
G = First guess at initial slope
G2 = Second guess at initial slope
RI = Final result at endpoint (using GI)
R2 = Second result at endpoint (using G2)
D = the desired value at the endpoint
Finite Difference method
- - +
dy
=

dx

yi+I - yI_I
2Ax

AX

i_p1

Jy11

_(2_zx2q1)y1

+1i+piJyi+i

Thomas Algorithm
a. = d1 c118_ 1

a1 = d1

wI=

cx.
WI -

a.

a1

10

=&2

You might also like