The Schur-Cohn Algorithm

Similar documents
p (i.e., the set of all nonnegative real numbers). Similarly, Z will denote the set of all

Chapter Newton-Raphson Method of Solving a Nonlinear Equation

Rank One Update And the Google Matrix by Al Bernstein Signal Science, LLC

GAUSS ELIMINATION. Consider the following system of algebraic linear equations

Effects of polarization on the reflected wave

Chapter Newton-Raphson Method of Solving a Nonlinear Equation

Remember: Project Proposals are due April 11.

The Number of Rows which Equal Certain Row

UNIVERSITY OF IOANNINA DEPARTMENT OF ECONOMICS. M.Sc. in Economics MICROECONOMIC THEORY I. Problem Set II

Model Fitting and Robust Regression Methods

Lecture 4: Piecewise Cubic Interpolation

Proof that if Voting is Perfect in One Dimension, then the First. Eigenvector Extracted from the Double-Centered Transformed

Variable time amplitude amplification and quantum algorithms for linear algebra. Andris Ambainis University of Latvia

4. Eccentric axial loading, cross-section core

ψ ij has the eigenvalue

DCDM BUSINESS SCHOOL NUMERICAL METHODS (COS 233-8) Solutions to Assignment 3. x f(x)

Dennis Bricker, 2001 Dept of Industrial Engineering The University of Iowa. MDP: Taxi page 1

Math 497C Sep 17, Curves and Surfaces Fall 2004, PSU

Review of linear algebra. Nuno Vasconcelos UCSD

Jens Siebel (University of Applied Sciences Kaiserslautern) An Interactive Introduction to Complex Numbers

MSC: Primary 11A15, Secondary 11A07, 11E25 Keywords: Reciprocity law; octic residue; congruence; quartic Jacobi symbol

COMPLEX NUMBER & QUADRATIC EQUATION

Chapter 2 Introduction to Algebra. Dr. Chih-Peng Li ( 李 )

Many-Body Calculations of the Isotope Shift

Two Coefficients of the Dyson Product

CISE 301: Numerical Methods Lecture 5, Topic 4 Least Squares, Curve Fitting

Principle Component Analysis

Not-for-Publication Appendix to Optimal Asymptotic Least Aquares Estimation in a Singular Set-up

Katholieke Universiteit Leuven Department of Computer Science

Quiz: Experimental Physics Lab-I

Partially Observable Systems. 1 Partially Observable Markov Decision Process (POMDP) Formalism

INTRODUCTION TO COMPLEX NUMBERS

Moment estimates for chaoses generated by symmetric random variables with logarithmically convex tails

Mechanical resonance theory and applications

Demand. Demand and Comparative Statics. Graphically. Marshallian Demand. ECON 370: Microeconomic Theory Summer 2004 Rice University Stanley Gilbert

ANALOG CIRCUIT SIMULATION BY STATE VARIABLE METHOD

8. INVERSE Z-TRANSFORM

DESIGN OF MULTILOOP CONTROLLER FOR THREE TANK PROCESS USING CDM TECHNIQUES

Applied Statistics Qualifier Examination

Non-Ideality Through Fugacity and Activity

Electrochemical Thermodynamics. Interfaces and Energy Conversion

HAMILTON-JACOBI TREATMENT OF LAGRANGIAN WITH FERMIONIC AND SCALAR FIELD

6 Roots of Equations: Open Methods

Lecture 36. Finite Element Methods

Research Article On the Upper Bounds of Eigenvalues for a Class of Systems of Ordinary Differential Equations with Higher Order

LAPLACE TRANSFORM SOLUTION OF THE PROBLEM OF TIME-FRACTIONAL HEAT CONDUCTION IN A TWO-LAYERED SLAB

Chapter I Vector Analysis

Introduction to Numerical Integration Part II

Study of Trapezoidal Fuzzy Linear System of Equations S. M. Bargir 1, *, M. S. Bapat 2, J. D. Yadav 3 1

Digital PI Controller Equations

Overview. Regular Languages. Finite Automata. A finite automaton. Startstate : q Acceptstate : q. Transitions

II The Z Transform. Topics to be covered. 1. Introduction. 2. The Z transform. 3. Z transforms of elementary functions

Mixture of Gaussians Expectation Maximization (EM) Part 2

Numerical Solution of Freholm-Volterra Integral Equations by Using Scaling Function Interpolation Method

13 Design of Revetments, Seawalls and Bulkheads Forces & Earth Pressures

ESCI 342 Atmospheric Dynamics I Lesson 1 Vectors and Vector Calculus

Duke Math Meet

Triangle-based Consistencies for Cost Function Networks

A Tri-Valued Belief Network Model for Information Retrieval

Introduction to Group Theory Note 2 Theory of Representation

Name: SID: Discussion Session:

Dr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur

FINITE NEUTROSOPHIC COMPLEX NUMBERS. W. B. Vasantha Kandasamy Florentin Smarandache

AQA Further Pure 1. Complex Numbers. Section 1: Introduction to Complex Numbers. The number system

International Journal of Pure and Applied Sciences and Technology

LOCAL FRACTIONAL LAPLACE SERIES EXPANSION METHOD FOR DIFFUSION EQUATION ARISING IN FRACTAL HEAT TRANSFER

THE COMBINED SHEPARD ABEL GONCHAROV UNIVARIATE OPERATOR

Numbers Related to Bernoulli-Goss Numbers

( ) 2 ( ) ( ) Problem Set 4 Suggested Solutions. Problem 1

COMPLEX NUMBERS INDEX

Solubilities and Thermodynamic Properties of SO 2 in Ionic

Digital Signal Processing

PRIMES AND QUADRATIC RECIPROCITY

Computing a complete histogram of an image in Log(n) steps and minimum expected memory requirements using hypercubes

Pyramid Algorithms for Barycentric Rational Interpolation

Abhilasha Classes Class- XII Date: SOLUTION (Chap - 9,10,12) MM 50 Mob no

Linear and Nonlinear Optimization

Announcements. Image Formation: Outline. The course. Image Formation and Cameras (cont.)

Definition of Tracking

Supplement 4 Permutations, Legendre symbol and quadratic reciprocity

MATHEMATICAL MODEL AND STATISTICAL ANALYSIS OF THE TENSILE STRENGTH (Rm) OF THE STEEL QUALITY J55 API 5CT BEFORE AND AFTER THE FORMING OF THE PIPES

Least squares. Václav Hlaváč. Czech Technical University in Prague

Chapter Runge-Kutta 2nd Order Method for Ordinary Differential Equations

M/G/1/GD/ / System. ! Pollaczek-Khinchin (PK) Equation. ! Steady-state probabilities. ! Finding L, W q, W. ! π 0 = 1 ρ

Lecture 3. Ax x i a i. i i

( s) Use of Transformations in Active BP Filter Designs

CENTROID (AĞIRLIK MERKEZİ )

Lumpability and Absorbing Markov Chains

Supplementary Material for Spectral Clustering based on the graph p-laplacian

Lecture 3 ( ) (translated and slightly adapted from lecture notes by Martin Klazar)

Grover s Algorithm + Quantum Zeno Effect + Vaidman

The Dirac Equation for a One-electron atom. In this section we will derive the Dirac equation for a one-electron atom.

Cramer-Rao Lower Bound for a Nonlinear Filtering Problem with Multiplicative Measurement Errors and Forcing Noise

Logistic regression with one predictor. STK4900/ Lecture 7. Program

Online Appendix to. Mandating Behavioral Conformity in Social Groups with Conformist Members

E-Companion: Mathematical Proofs

COMP 465: Data Mining More on PageRank

Multiple view geometry

Structure and Drive Paul A. Jensen Copyright July 20, 2003

Patterns of Continued Fractions with a Positive Integer as a Gap

Transcription:

Modelng, Estmton nd Otml Flterng n Sgnl Processng Mohmed Njm Coyrght 8, ISTE Ltd. Aendx F The Schur-Cohn Algorthm In ths endx, our m s to resent the Schur-Cohn lgorthm [] whch s often used s crteron for testng the stblty of bounded-nut bounded-outut systems []. To smlfy the descrton of ths lgorthm, we frst te u the nlyss of the stblty domn of nd -order trnsfer functon. Ths rtculr cse leds to smlfcton of the stblty crter mosed on the denomntor of the trnsfer functon. Unfortuntely, t cnnot be led to trnsfer functons of n order greter thn. We lso resent the Schur-Cohn stblty lgorthm bsed on the trnsfer functon of n ll-ss flter, llowng us to estblsh equvlence relton between the Schur coeffcents nd the reflecton coeffcents. Let there be second-order trnsfer functon defned s follows: b b N [F.] The oles of b b re equl to: 4 nd 4 [F.]

36 Modelng, Estmton nd Otml Flterng n Sgnl Processng nd ts eros re defned s follows: b b 4b nd b b 4b [F.3] eendng on the vlues ten by nd, the oles cn be rel or comlex. For exmle, when 4, the oles re comlex conjugtes of ech other. Otherwse, they re rel. To ensure stblty, the oles of the trnsfer functon must be locted wthn the unt crcle n the -lne,.e.: nd [F.4] Ths constrnt mles tht the followng two nequltes re stsfed: [F.5] nd: [F.6] Reltons [F.5] nd [F.6] me t ossble to defne trngle n the, lne where the flter s stble nd whch s clled the stblty trngle. Ths trngle dected n Fgure F.. s smle tool for testng the stblty s t s bsed on the vlues of the flter s coeffcents.

Aendces 363 Fgure F.. The stblty trngle Alcton of the stblty trngle Let there be th -order trnsfer functon defned s follows: N [F.7] where. The frst condton requred for the stblty s exressed n terms of : snce,.., [F.8] In the rest of ths stblty test, we wll te nd ssume tht the frst condton [F.8] s stsfed. Let us develo the trnsfer functon of th -order ll-ss flter usng. [F.9]

364 Modelng, Estmton nd Otml Flterng n Sgnl Processng Furthermore, we defne s follows:...... [F.] Note : We note tht s lso n ll-ss flter of order -, nd ts exresson cn be smlfed by mosng:,.., [F.] Thus, we hve:...... [F.] From equton [F.], we note tht the oles of,, re such tht: [F.3] Thus, tng equton [F.8] nto ccount, we obtn: [F.4] We now show tht stsfyng the ssertons s the trnsfer functon of stble ll-ss flter nd s equvlent to syng tht s stble. To do ths, we frst show tht f s the trnsfer functon of stble llss flter nd, then s the trnsfer functon of stble ll-ss flter.

We cn esly show tht ny gven ll-ss functon roertes: G G G f f f Aendces 365 G stsfes the followng [F.5] Consequently, f s n fct the trnsfer functon of stble ll-ss flter, when. owever, from equton [F.4], we see tht. Therefore, the oles of le nsde the unt crcle n the -lne nd s stble.. Let us te u equton [F.] nd exress Let us now ssume tht nd tht. We thus obtn: If s ole of s the trnsfer functon of stble ll-ss flter s functon of [F.6], t must verfy: [F.7] As, we get: [F.8] Snce gves us: s the trnsfer functon of stble ll-ss flter, equton [F.5] [F.9] f

366 Modelng, Estmton nd Otml Flterng n Sgnl Processng [F.] f [F.] f Condton [F.8] contrdcts [F.9], but s n greement wth [F.]. Consequently, s stble. nd Usng the sme develoment s resented bove, we cn defne n terms of, then defne 3 n terms of, nd so on, untl we obtn. At ech successve ste, we test the vlue of, then, nd so on. The Schur stblty crteron sttes tht s stble f for ll vlues of. Let us now loo t the corresondence between Schur coeffcents nd reflecton coeffcents. Let us te equton [F.]:,.., P [F.] nd wrte t n mtrx form, tng nto ccount tht nd : [F.] Let us comre ths mtrx wth equton [.9], obtned usng the Levnson lgorthm, nto whch we ntegrte :

Aendces 367 [F.3] The followng olynoml s ssocted wth the vector T : Thus, we cn ssocte the followng olynoml wth the vector T : The Schur-Cohn lgorthm s wrtten s follows: [F.4] whle the exresson for the Levnson lgorthm stsfes: [F.5] We note tht the mtrces n equtons [F.4] nd [F.5] re the nverses of one nother. Thus, we hve:

368 Modelng, Estmton nd Otml Flterng n Sgnl Processng [F.6] so: [F.7] Ths roves the equvlence between the Schur coeffcents nd the reflecton coeffcents. References [] T. Klth, A Theorem of I. Schur nd ts Imct on Modern Sgnl Processng, Oertor Theory: Advnces nd Alctons I. Schur Methods n Oertor Theory nd Sgnl Processng, 8,. 9-3, Brhuser, 986. [] S. K. Mtr, gtl Sgnl Processng Comuter Bsed Aroch, 3 rd edton, McGrw-ll, 6.