PSEUDOSPECTRA o Application of eigenvalue o Pseudospectra definition - - PowerPoint PPT Presentation

pseudospectra
SMART_READER_LITE
LIVE PREVIEW

PSEUDOSPECTRA o Application of eigenvalue o Pseudospectra definition - - PowerPoint PPT Presentation

29/09/1438 OVERVI EW PSEUDOSPECTRA o Application of eigenvalue o Pseudospectra definition D E F I N I T I O N S A N D o Examples of pseudospectra A P P L I C A T I O N S o Normal matrices o Ill-condition problems S U P E R V I S O R D R . N


slide-1
SLIDE 1

29/09/1438 1

PSEUDOSPECTRA

D E F I N I T I O N S A N D A P P L I C A T I O N S

S U P E R V I S O R D R . N O B A K H T I

H A M I D Z A R G A R A N

OVERVI EW

  • Application of eigenvalue
  • Pseudospectra definition
  • Examples of pseudospectra
  • Normal matrices
  • Ill-condition problems
  • Fragility of controllers

2/21

W HY EI GEN VALUES?

Diagonalization and separation of variables Resonance: heightened response to selected inputs. Asymptotics and stability: dominant response to general inputs They give a matrix a personality

3/21

QUI Z

1 1 1 , 1 5 2

Which curve is which?

4/21

slide-2
SLIDE 2

29/09/1438 2

PSEUDOSPECTRA TOOL

Better question is … Is large? Is z an eigenvalue of A? Is singular?

5/21

  • First def.

PSEUDOSPECTRA DEFI NI TI ONS

  • Example:

1

6/21

  • Second def.

∈ ∈ ∈ .

PSEUDOSPECTRA DEFI NI TI ONS

( CONTI NUED)

7/21

  • Third def.

∈ ∈ 1.

The three definitions are equivalent

PSEUDOSPECTRA DEFI NI TI ONS

( CONTI NUED)

8/21

slide-3
SLIDE 3

29/09/1438 3

EXAM PLES

1 1 1 1 5 2

9/21

NORM AL M ATRI CES

  • Normal matrix

↔ ∗ ∗

  • Some other equivalent conditions:
  • ,

10/21

NORM AL M ATRI CES

  • In general we have :

| , ⊆ If A is normal then = | , And for 2-norm ∆

11/21

EXAM PLE

1 1

  • B is a normal matrix

12/21

slide-4
SLIDE 4

29/09/1438 4

APPLI CATI ON I N TRANSI ENT RESPONSE ANALYSI S

  • Theorem.

For all 0 we have sup

  • sup
  • and

sup lim

1 log

13/21

I LL-CONDI TI ON PROBLEM S

  • consider a polynomial with zeros

2, 2, … , 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

Zeros of 2

14/21

I LL-CONDI TI ON PROBLEM S

1 20 ⋯ 2 20 ⋯ ⋮ 3 ⋯ ⋮ ⋮ ⋱ 20

20

  • Characteristic equation:

20

  • 20

20 ! 1 !

  • 15/21

FRAGI LI TY OF CONTROLLERS

 Any controller should be able to tolerate some uncertainty

in its coefficients

 Inherent imprecision in analog-digital and digital-analog conversion  Finite resolution measuring instruments  Round off errors in numerical computations

16/21

slide-5
SLIDE 5

29/09/1438 5

FRAGI LI TY OF CONTROLLERS

EXAM PLE

  • Electromagnetic suspension system
  • µ-synthesis technique

Design controller

17/21

FRAGI LI TY OF CONTROLLERS

EXAM PLE( CONTI NUED)

  • The poles of closed-loop system
  • The normalized ratio of required change in controller coefficients to

destabilize the closed loop

  • ‖0‖ 1.455 352 715 525 003 1015

18/21

W HY PSEUDOSPECTRA COULD BE A SOLUTI ON?

Σ Perturbed MIMO system Designed MIMO controller

19/21

REFRENCES

  • L. N. Trefethen, M. Embree, Spectra and pseudospectra: the behavior of nonnormal matrices

and operators, New Jersey, Princeton University Press, 2005, pp. 12-21

  • J. H. Wilkinson, Rounding Errors in Algebraic Processes, New York, Dover Publications, pp

44-45 & 138-139.

  • L. H. Keel, P. Bhattacharyya, Robust, fragile or optimal?, American Control Conference, 1997.

Proceedings of the 1997. Vol. 2. IEEE, 1997.

  • A. Jensen, Lecture Notes on Spectra and Pseudospectra of Matrices and Operators,

Department of Mathematical Sciences, Aalborg University, 2009.

20/21

slide-6
SLIDE 6

29/09/1438 6

THANKS FOR YOUR ATTENTI ON ANY QUESTI ON?

21/21