NATIONAL CHENG KUNG UNIVERSITY

Department of Mechanical Engineering

LINEAR SYSTEM HOMEWORK 2

Instructor: Student: Student ID: Department: Class:

Prof. Szu Chi Tien Nguyen Van Thanh P96007019 Inst. of Manufacturing & Information Systems 1001- N154000 Linear System

October 10, 2011

Contents
Problem 1 ............................................................................................................................... 2 Problem 2 ............................................................................................................................. 13 Problem 3 ............................................................................................................................. 18

Problem 1
Consider the following linear system:

1. Find the eigenvalues and eigenvectors of the system matrix A (you may use Matlab). Is the system stable? Explain. The characteristic polynomial equation of the matrix A is, ( ) .0 { The eigenvectors of A is defined by an expression: ( ) 0 10 1 { 1/ ( )

- With {

: ( ( : { ( ( ) ) [ ] ) ) [ ]

- With

The system is stable since al the eigenvalues of matrix A those real part are negative.

2. Find the modal matrix Am = T

AT , where columns of the similarity transformation

matrix T are the eigenvectors of the matrix A. Is the modal matrix unique? Explain. , [ 0 ] 1[ ] ]

The modal matrix is not unique. It depends on how we specify the eigenvectors in the similarity transformation matrix T. For example, if we change T: , [ ] 0 1[ ] ]

Note: one eigenvalue can have more than one eigenvector. Now, we change , and but still keep their position in the similarity transformation matrix T. For example, , So Am doesnt change. That means: , , , , -

3. Find the exponential matrix e using four different methods: (a) Cayley Hamilton theorem (Finite series representation). We can express the exponential matrix ( Where the coefficients ( ) and ( ) ) as a form: ( ) (3a.1)

At

( ) are solutions of the differential equations [ { ( ) ] ( ) 0 1[ 0 1 ( ) ] ( )

( ) [ ] ( )

(3a.2)

By using the Matlab commands: >> syms t; >> S = dsolve('Dalpha0 = -4*alpha1',... 'Dalpha1 = alpha0 - 2*alpha1',... 'alpha0(0) = 1, alpha1(0) = 0'); >> alpha0 = S.alpha0; >> alpha1 = S.alpha1; >> pretty(alpha0) >> pretty(alpha1)

We obtain the solution: ( ) ( ( ) ( )

( )) (3a.3)

( )

Another way for obtaining

( ) and

( ) is using the expression below. ( ) (3a.4)

Where

are the eigenvalues of matrix A. 4

( ) ( )

( ) ( )

( ) ( ) (3a.5)

From section 1, we already known the eigenvectors of matrix A, they are: { By using the Eulers formula: ( )
(

(3a.6)

( ) ( ( ( ) ( ) ( )) ( ))

(3a.7) (3a.8)

) )

Now, we plug equations (1.2) and (1.3) into equation (1.1), we obtain. ( ) ( ( ) ( )

( )) (3a.9)

( )

Plug equation (3a.3) (or (3a.9)), ( ) ( ) ( ( [ ( ) ( ) ( )) 0 ( )) ( 1 ( ) ( )0 ( ) ( )) ] 1

( )

(b) Resolvent matrix (Inverse Laplace transform of (sI A) ). ( ) ( ) ( ( ) ) ( ) ( ) 5

Where, ( ) Use the Leverrier-Fadeev Algorithm, we obtain ( )

10 (

1 ) ( ) 1

{ ( ) ( ) ( 0 1 0

1 )

( )

( ) ( )

Taking the inverse Laplace transform of we obtain (

[ . /

( ) (by looking up the table of Laplace transform)

( )) (

. /

( )

( )

( ))

This can be verified by using the Matlab commands: syms s t; Phi11 = (s+2)/(s^2+2*s+4); Phi12 = 1/(s^2+2*s+4); Phi21 = -4/(s^2+2*s+4); Phi22 = s/(s^2+2*s+4); 6

Phi = [Phi11 Phi12; Phi21 Phi22]; pretty(ilaplace(Phi)) (c) Modal transformation T where . [
( ) ( )

( (

][

][

) )

] [

(
[

. /

( )) (

. /

( )

( )

( ))

(d) Use Maple or Mathematica (if you have accessed and familiar with one of these tools, note: this part is optional). In Maple, by using the commands:

4. Find the limit of the exponential matrix

as t .

( ([

( )

( )) (

( )

( ) ( )) ])

( )

Note that, all two eigenvalues of matrix A has negative real part the system is modeled by equation (1) is stable.

. Since,

5. Solve analytically for the time responses x(t) to initial conditions x(0) = [0 3] T and no applied input u(t)=0 for all t 0. Plot your time responses. 8

Time responses x(t) to initial conditions x(0) = [0 3]T and no applied input u(t)=0 for all t 0 are given by: ( ) ( ) ( [ ( ) ( ) [ ( )) ( ( ) ( ) ( ) 0 1 ( )) ]

( ) ( ( )

] ( ))

Time responses are shown in figure 5.1.

Figure 5.1 Time responses to initial conditions ( )

6. (a) Solve analytically for the time responses x(t) to zero initial conditions x(0) = [0 0] and with input u(t)=7 for all t 0.

What are the steady-state values of x(t); i.e.

( )?

Time responses x(t) to the initial conditions x(0) = [0 0] T and with the input u(t) = 7 for all t >= 0 are given by ( )
( )

( )
( )

( )

((
(

) )
)

(( ) )

) )/
( )

((

) ) ((

((

(( ) ) ((

) )

]0 1 ) )/

( )

[ ( )
( )

( ((

((

) )

) )) ]

[ (

( )

( ) ( )) ]

Where,

. By using the Hint or Maple, we obtain: ( ( ) [ ( ) ] ( ) ( ))

10

( ( ) [ Time response are shown in figure 6a.1.

( )

( )) [ ] ]

( )

Figure 6a.1 Time response to input u(t) = 7 (b) Find the steady-values using the state-space model given in equation (1) without solving for the time responses In the steady-state, we have 0 10 1 [ ]

11

7. Solve analytically for the time responses x(t) to initial conditions in Part (5) and at the same time with the applied input in Part (6). Plot your time responses. The system is a LTI, so by applying the properties of the LTI, we use the superposition property, ( ) ( ( ) ( ) ( ) ( ) ( ] ( )) [ ( ) ( ) ( )) ]

( )

( )

Time responses are shown in figure 7.1.

Figure 7.1 Time responses to x(0) = [0 3]T, and u(t) = 7 12

Problem 2
[ ] , -

Given the following matrix

1. (a) Find a set of orthonormal basis vectors for the range of A. Hint: Use Gram Schmidt orthogonalization. Recall that, if We have, ( ) ( ) , the y is the range space of A, where x is any vector in R3.

From matrix A, the first column of matrix A = the third column the second column of matrix A, that means

Hence, rank(A) = 2 (or we can using the Matlab command: >> rank(A)). So, the basis vectors for the range of A ( ) are * + (because and are linearly independent). Now, we use Gram-Schmidt orthogonalization to orthogonalize the * +. | | || ] || || [ ]

[ ]

[ ]

|| ||

A set of orthonormal basis vectors for the range space of A is * + (b) Can you find a solution x R such that
3

13

From

, we can find a solution

iff

. However, , the solution x is not unique. * +, where,

the matrix A is not full rank, if

We choose a set of basis vectors for the range space of matrix A is

[ ]

How we know is in the range space of A or not? If we can find two coefficients such that,

are not simultaneously zero. Hence, is in the range space of A. If we cannot find satisfying (i) Find , satisfying, . Hence, is not in the range space of A.

[ ]

Hence, we can find a solution

. Because, the matrix A is not full

such that

rank, so the solution x is not unique. (ii) Find , satisfying,

14

[ ]

Hence, we cannot find a solution

such that

2. Find a set of orthonormal basis vectors for the null space of A. Recall, then is in null space of A.

][ ]

][ ]

Thus, the null space of A is spanned by the vector

Hence, an orthonormal basis vector for the null space of A is || || ]

3. Use MATLAB to determine the singular value decomposition of A. Singular value decomposition:

Use the command [U, S, V] = svd(A) in Matlab, we obtain

15

4. Using the results from the singular value decomposition in Part 3, answer the following questions: (a) What are the singular values (1, 2, 3) of A? Note that 1 2 3.

(b) What is the rank of A? Verify your result using MATLAB rank(A). ( ) ( )

Use the command rank(A) in Matlab (>> rank(A)), we obtain the ans = 2. (c) What is the right singular vector associated with the singular value 3? Show the connection of the right singular vector with the null space of A in Part 2. , -, where, are the right singular vectors of A. The right singular

vector associated with the singular value 3 is

Recall from part 2, the null space of A is spanned by the vector

16

it is along the same direction as the null space of A in Part 2.

(d) What are the left singular vectors corresponding to the non-zero singular values of A? Show the connection of the left singular vectors with the range space of A in Part 1. , -, where, are the left singular vectors of A. The left singular

vectors corresponding to the non-zero singular values of A is

We choose a set of basis vectors for the range space of matrix A is

* +, where,

[ ]

We have, { Hence,

are in the range space of A in Part 1. We can verify the result by using

the command rank([u1 u2 U1 U2]) = 2.

17

Problem 3
[ ]

Given the following matrix

1. Find the eigenvalues and eigenvectors of the matrix ATA using Matlab. The eigenvector matrix of ATA forms the unitary matrix V in the singular value decomposition of A according to equation (5). By using the following commands in Matlab, >> A = [1 2; 2 7; 4 0]; >> [V1, D1] = eig(A'*A) We obtain 0 So, the eigenvalues of the matrix ATA are corresponding to two eigenvectors 0
T

1 ( )

0 ( 1 0 1

1 )

2. Find the eigenvalues and eigenvectors of AA using MATLAB. The eigenvector matrix
T

of AA forms the unitary matrix U in the singular value decomposition of A according to equation (8). By using the following commands in Matlab, >> A = [1 2; 2 7; 4 0]; >> [V2, D2] = eig(A*A') We obtain 18

So, the eigenvalues of the matrix AAT are

(exactly the same to the eigenvalues of the matrix ATA) corresponding to two eigenvectors [ ] [ ]

3. Finally, determine the singular values i (i =1, 2 r m) of the matrix A from the square root of the eigenvalues of A A. By using the following commands in Matlab, >> A = [1 2; 2 7; 4 0]; >> [V1, D1] = eig(A'*A); >> SigmaMatrix = D1^(1/2) We obtain [ ] 0 1
T

4. Compare the results obtained from the Matlab command svd(A) to the unitary matrices U and V obtained in Parts (1) and (2). By using the following commands in Matlab, >> A = [1 2; 2 7; 4 0]; >> [U, S, V] = svd(A) We obtain 19

Recall, from Part 1: 0 And from Part2: 1 0 1

Thus, the left singular vectors in U are the same as the eigenvectors in Part2 and the right singular vectors V are the same as the eigenvectors in Part1. If we specify the eigenvalues in given order (i.e. descending), hence, {

20