Вы находитесь на странице: 1из 7

MATRIX

Transpose of a matrix: Change of row and column


Symmetric matrix
Skew-symmetric matrix
Identity matrix
 M 11 M 21 " M N1  {a}T  {a1 a2 " aN }
M M 22 " MN2 
[M ]   12
T 
 # # % #  1 0 0 
 
 M 1K M 2K " M NK  [I ]   0 1 0 
 
0 0 1 
 M 11 M 12 " M 1N 
M M 22 " M 2 N 
[S]  [S]   12
T  [ A]T  [ A]
 # # % # 
 
 M 1N M 2 N " M NN 

VECTOR-MATRIX CALCULUS
Addition
{c}  {a}  {b},  ci  ai  bi , i  1,", N
{d}  {a}  {b},  di  ai  bi , i  1,", N
k{a}  {ka1 ka2 " kaN }T
[C]  [ A]  [B],  Cij  Aij  Bij , i  1,", N , j  1,", K
[D]  [ A]  [B],  Dij  Aij  Bij , i  1,", N , j  1,", K
Scalar product
a1
b1

a  b  a1b1  a2b2  a3b3


a  a2 and b  b2
a  b  a b cos a b
 3  3

Norm (Magnitude of a vector)


a  a a

4
VECTOR-MATRIX CALCULUS cont.
Vector product
i j k
a  b  a1 a2 a3
b1 b2 b3
 (a2b3  a3b2 )i  (a3b1  a1b3 ) j  (a1b2  a2b1 )k
a2b3  a3b2


 a3b1  a1b3
a b  a b
 1 2 2 1 ab
b
a  b  a b sin  n n
aa  0

b  a  a  b a

MATRIX-VECTOR MULTIPLICATION
Matrix  Vector = Vector
 m11 m12 m13  a1
m11a1  m12 a2  m13a3


c  [M ]  a   m21 m22 m23   a2  m21a1  m22 a2  m23a3
 
 m31 m32 m33   a3   m31a1  m32 a2  m33a3 
3
ci   M ij a j , i  1, 2,3
j 1

Vector  Matrix  Vector = Scalar


b  [M ]  a  b1 (m11a1  m12 a2  m13a3 )
b2 (m21a1  m22 a2  m23a3 )
b3 (m31a1  m32 a2  m33a3 )

6
MATRIX-MATRIX MULTIPLICATION
Matrix  Matrix = Matrix
3
[C]  [ A][B] CIJ   AIK BKJ , I  1, 2,3, J  1, 2,3
K 1

Determinant
Matrix is singular when its determinant is zero

a11 a12 a11 a12 a13


A  a11a22  a12 a21
a21 a22 A  a21 a22 a23
a31 a32 a33
1
Inverse: [ A][ A]  [I]  a11 (a22 a33  a23a32 )
1  a12 (a21a33  a23a31 )
a a12  1  a22 a12 
[ A]1   11   a13 (a21a32  a22 a31 )
 a21 a22  A  a21 a11 

RULES OF MATRIX MULTIPLICATION

Associative rule: (AB)C  A(BC)

Distributive rule: A(B  C)  AB  AC

Non-commutative: AB  BA

Transpose of product: ( AB)T  BT AT , (ABC)T  CT BT AT

Inverse of product: ( AB)1  B1A1 , ( ABC)1  C1B1A1

8
MATRIX EQUATION
N unknowns (x1, x2, , xN) and N equations
unique solution if all equations are independent

a11 x1  a12 x2  "  a1N xN  b1


a21 x1  a22 x2  "  a2 N xN  b2  a11 a12 " a1N 
# a a22 " a2 N 
[ A]   21 
aN 1 x1  aN 2 x2  "  aNN xN  bN  # # % # 
 
 aN 1 aN 2 " aNN 
Matrix form:

[ A] {x}  {b}
x1
b1

Solution: [A]1 exists or [A] is not singular x b


2
{x}  , {b}  2
[ A]1[ A] {x}  [ A]1 {b} # #
[I ] {x}  [ A]1 {b}  xN  bN 
{x}  [ A]1 {b}
9

EIGEN VALUE AND EIGEN VECTOR


Eigen value problem
[ A] {x}  {x}
 : Eigen value
{x} : Eigen vector
How to solve?
[ A]  {x}  {x}  {0}
[ A]  {x}  [I]{x}  {0}
[ A   I ]  {x}  {0}
In order to have non-trivial solution, the determinant must be zero.
A  I  0
Calculate  from this equation and calculate {x} from the eigen value
problem

10
QUADRATIC FORM
Quadratic form: quadratic function of all components

J  a11 x12  a22 x22  "  ann xn2


 a12 x1 x2  a13 x1 x3  "  an,n1 xn xn1

Matrix notation  a11 a12 " a1n  x1

a a22 " a2 n  x2 n n
J  {x} [ A]{x}  {x1 , x2 ,!, xn }  21
T    a x x
 # # % #  # i 1 j 1 ij i j
 
 an1 an 2 " ann   xn 
Symmetric part is enough

J  {x}T [B]{x} J *  {x}T [B* ]{x} [ B* ] 


1
2
 [B]  [B]T 

J  {x}T [B]{x}  {x}T [B* ]{x}  J *

11

POSITIVE DEFINITE MATRIX


Positive definite

{x}T [ A]{x}  0, for all {x} in R n


{x}T [ A]{x}  0, only if {x}  {0}

Positive semi-definite

{x}T [ A]{x}  0, for all {x} in R n


{x}T [ A]{x}  0, for some {x}  {0}

Positive definiteness = each column of the matrix is linearly


independent = the matrix is invertible.

12
MAXIMA & MINIMA OF FUNCTIONS
Single Variable f ( x)
Taylor series expansion

f 1 2 f
f ( x  x)  f ( x )  x  (x) 2  "
x x x 2 x 2 x x

In order for f to be extremum,

f
0
x x x
2 f
Condition for minima: 0
x 2 x x

2 f
Condition for maxima: 0
x 2 x x

13

MAXIMA & MINIMA OF FUNCTIONS cont.


Multi-Variable f (x)
Taylor series expansion
n
f 1 n n 2 f
f ( x  x)  f ( x )   xi   xi x j  "
i 1 xi x x
2 i 1 j 1 xi x j
x x
In order for f to be extremum, Hessian
matrix Hij
f f f
 "  0
x1 x2 xn
Condition for minima: [H] is positive definite
Condition for maxima: [H] is negative definite

14
MINIMUM PRINCIPLE
Function in quadratic form

F  {x}T [ A]{x}  2{x}T {c}

[A]: stiffness of the structure, {x}: displacement, {c}: applied force


twice of the potential energy
structure is in equilibrium when F has a minimum value

Matrix equation
[ A]{x}  {c}

Solution of the matrix equation minimizes the quadratic form F.

15

Courtesy of Nam-Ho Kim

Вам также может понравиться