MEM 255 Introduction to Control Systems Review: Basics of Linear Algebra

Similar documents
Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix

Formulas for the Determinant

APPENDIX A Some Linear Algebra

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0

332600_08_1.qxp 4/17/08 11:29 AM Page 481

= = = (a) Use the MATLAB command rref to solve the system. (b) Let A be the coefficient matrix and B be the right-hand side of the system.

Norms, Condition Numbers, Eigenvalues and Eigenvectors

MATH Homework #2

COMPLEX NUMBERS AND QUADRATIC EQUATIONS

2 More examples with details

Salmon: Lectures on partial differential equations. Consider the general linear, second-order PDE in the form. ,x 2

Section 8.3 Polar Form of Complex Numbers

Quantum Mechanics I - Session 4

7. Products and matrix elements

2.3 Nilpotent endomorphisms

Linear Algebra and its Applications

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems

Quantum Mechanics for Scientists and Engineers. David Miller

DISCRIMINANTS AND RAMIFIED PRIMES. 1. Introduction A prime number p is said to be ramified in a number field K if the prime ideal factorization

Bezier curves. Michael S. Floater. August 25, These notes provide an introduction to Bezier curves. i=0

763622S ADVANCED QUANTUM MECHANICS Solution Set 1 Spring c n a n. c n 2 = 1.

Transfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system

The Jacobsthal and Jacobsthal-Lucas Numbers via Square Roots of Matrices

The Order Relation and Trace Inequalities for. Hermitian Operators

Number of cases Number of factors Number of covariates Number of levels of factor i. Value of the dependent variable for case k

Example: (13320, 22140) =? Solution #1: The divisors of are 1, 2, 3, 4, 5, 6, 9, 10, 12, 15, 18, 20, 27, 30, 36, 41,

Fall 2012 Analysis of Experimental Measurements B. Eisenstein/rev. S. Errede

SL n (F ) Equals its Own Derived Group

Feb 14: Spatial analysis of data fields

Lecture 6/7 (February 10/12, 2014) DIRAC EQUATION. The non-relativistic Schrödinger equation was obtained by noting that the Hamiltonian 2

Advanced Quantum Mechanics

Time-Varying Systems and Computations Lecture 6

Bézier curves. Michael S. Floater. September 10, These notes provide an introduction to Bézier curves. i=0

Problem Do any of the following determine homomorphisms from GL n (C) to GL n (C)?

Singular Value Decomposition: Theory and Applications

SUMMARY OF STOICHIOMETRIC RELATIONS AND MEASURE OF REACTIONS' PROGRESS AND COMPOSITION FOR MULTIPLE REACTIONS

5 The Rational Canonical Form

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence.

MEM633 Lectures 7&8. Chapter 4. Descriptions of MIMO Systems 4-1 Direct Realizations. (i) x u. y x

CHAPTER 4. Vector Spaces

Integrals and Invariants of Euler-Lagrange Equations

Homework Notes Week 7

Mathematical Preparations

THE SUMMATION NOTATION Ʃ

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method

Errors for Linear Systems

Chapter 7 Generalized and Weighted Least Squares Estimation. In this method, the deviation between the observed and expected values of

The exponential map of GL(N)

Section 3.6 Complex Zeros

Lecture Notes Introduction to Cluster Algebra

Workshop: Approximating energies and wave functions Quantum aspects of physical chemistry

Dynamic Programming. Preview. Dynamic Programming. Dynamic Programming. Dynamic Programming (Example: Fibonacci Sequence)

(A and B must have the same dmensons to be able to add them together.) Addton s commutatve and assocatve, just lke regular addton. A matrx A multpled

1 (1 + ( )) = 1 8 ( ) = (c) Carrying out the Taylor expansion, in this case, the series truncates at second order:

Randić Energy and Randić Estrada Index of a Graph

Perron Vectors of an Irreducible Nonnegative Interval Matrix

Unit 5: Quadratic Equations & Functions

THEOREMS OF QUANTUM MECHANICS

However, since P is a symmetric idempotent matrix, of P are either 0 or 1 [Eigen-values

12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA. 4. Tensor product

Algebraic Solutions to Multidimensional Minimax Location Problems with Chebyshev Distance

Composite Hypotheses testing

U.C. Berkeley CS294: Beyond Worst-Case Analysis Luca Trevisan September 5, 2017

p 1 c 2 + p 2 c 2 + p 3 c p m c 2

Lecture 3: Dual problems and Kernels

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography

A New Refinement of Jacobi Method for Solution of Linear System Equations AX=b

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 13

w ). Then use the Cauchy-Schwartz inequality ( v w v w ).] = in R 4. Can you find a vector u 4 in R 4 such that the

Lecture 3. Ax x i a i. i i

MMA and GCMMA two methods for nonlinear optimization

CHAPTER III Neural Networks as Associative Memory

Tensor Analysis. For orthogonal curvilinear coordinates, ˆ ˆ (98) Expanding the derivative, we have, ˆ. h q. . h q h q

MATH 241B FUNCTIONAL ANALYSIS - NOTES EXAMPLES OF C ALGEBRAS

THE CHINESE REMAINDER THEOREM. We should thank the Chinese for their wonderful remainder theorem. Glenn Stevens

8.6 The Complex Number System

CHAPTER 14 GENERAL PERTURBATION THEORY

9 Characteristic classes

Orthogonal Functions and Fourier Series. University of Texas at Austin CS384G - Computer Graphics Spring 2010 Don Fussell

Lecture 10 Support Vector Machines II

Fundamental loop-current method using virtual voltage sources technique for special cases

Affine transformations and convexity

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity

Statistical Mechanics and Combinatorics : Lecture III

BOUNDEDNESS OF THE RIESZ TRANSFORM WITH MATRIX A 2 WEIGHTS

An efficient algorithm for multivariate Maclaurin Newton transformation

Outline and Reading. Dynamic Programming. Dynamic Programming revealed. Computing Fibonacci. The General Dynamic Programming Technique

A new Approach for Solving Linear Ordinary Differential Equations

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )

Orthogonal Functions and Fourier Series. University of Texas at Austin CS384G - Computer Graphics Fall 2010 Don Fussell

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS)

Some Consequences. Example of Extended Euclidean Algorithm. The Fundamental Theorem of Arithmetic, II. Characterizing the GCD and LCM

Difference Equations

Deriving the X-Z Identity from Auxiliary Space Method

NOTES ON SIMPLIFICATION OF MATRICES

Math 217 Fall 2013 Homework 2 Solutions

Complex Numbers. x = B B 2 4AC 2A. or x = x = 2 ± 4 4 (1) (5) 2 (1)

The Degrees of Nilpotency of Nilpotent Derivations on the Ring of Matrices

Transcription:

MEM 255 Introducton to Control Systems Revew: Bascs of Lnear Algebra Harry G. Kwatny Department of Mechancal Engneerng & Mechancs Drexel Unversty

Outlne Vectors Matrces MATLAB Advanced Topcs

Vectors A vector s a one-dmensonal array of scalarelements (real or complex numbers) x1 x 2 column vector: x=, row vector: y = y1 y2 y xn Vectors of equal dmenson can be added (elementwse): x= a+ b x = a + b, = 1,, n 1 2 = Vectors can be multpled by a scalar: x= αa x = αa Vectors can be transposed: x x xn T [ x x x ] 1 2 n [ ] n

Inner Product & Norm The nner product of two n-dmensonal vectors x, y s xy, = xy, n = 1 for column vectors xy, = for row vectors xy, = xy The Eucldean norm or length of a vector x s x = x, x = x n = 1 2 T xy Other norms or length measures are also just useful: n x = x, x = max x 1 = 1 T

Lnear Combnatons of Vectors Suppose α, = 1,, p s a set of scalars and x, = 1,, p s a set of column or row vectors, then we defne a new vector y va the lnear combnaton: for columns y = α x + α x + + α x 1 1 2 2 y1 x1,1 x p,1 = α 1 + + α p yn x x 1 1 2 2 1, n p, n p p A set of p vectors x, = 1,, p s lnearly dependent f there exsts a nontrval set of constants α, = 1,, p such that α x + α x + + α x = 0 otherwse t s lnearly ndependent. p p A set of lnearly ndependent n-dmensonal vectors contans at most n vectors.

Matrces A matrx s a 2-dmensonal (rectangular) array of elements: m rows a11 a12 a1 n am1 amn n columns Sometmes we wrte A= aj The elements are called scalars, they are usually real or complex numbers. A matrx wth one row, m = 1, s called a row matrx or row vector. A matrx wth one column, n = 1, s called a column matrx or column vector.

Algebrac Operatons Equalty - two matrces (of the same sze) AB, are equal, wrtten B, f ther correspondng elements are equal, a = b, for1 m,1 j n j j Matrces of the same sze can be added and subtracted. Matrx A= addton and subtracton are performed element-wse A+ B= C a + b = c A B= C a b = c j j j j j j Any matrx A= aj can be multpled by a scalar α αa= αaj An m n matrx A can be post multpled by an n q matrx to produce an m q matrx C, = = C AB, c a b n j k= 1 k k B

Multplcaton b12 b 22 c 22 = a21 a22 a23 a 24 b23 b24 b12 b c a a a a a b [ ] 22 4 22 = 21 22 23 24 = k= 1 2k k2 b32 b 42

Transpose The transpose of an m n matrx s the n p matrx obtaned by nterchangng rows and columns: a11 am 1 a11 a12 a 1n a12 a m2 T A= A = am 1 am2 amn a1 n a mn T A matrx s symmetrc f A = A The followng rules obtan: ( AB) ( ) T T T = B A T T T A+ B = A + B

Determnant ( ) ( ) The j-th mnor M of a square n n matrx A s the n 1 n -1 submatrx of A obtaned by elmnatng row and column j. The determnant of a square matrx s defned recursvely. The determnant of a j= 1 j ( ) [ a ] 1 1 matrx s det = a. 11 11 The determnant of a n n matrx s defned by the expanson n det A= a γ for any = 1,2,, n j j where γ s the cofactor γ = 1 det M Note : 'for any ' means expand along any row. The same result s obtaned by expandng along any column. + j j j j

Propertes of Determnants multply any sngle row or column of A by scalar α to get A det A= α det A nterchange any two rows or columns of A to get A det A= det A add multple of any row or column to another row or column to get det A= det A T det A = det A, det AB= det Adet B for AC, square A B det det Adet D 0 D = for A nonsngular A B 1 det = det Adet D CA B C D A

Matrx Inverse An dentty matrx of sze n s the square matrx wth n rows and columns: 1 0 0 0 1 I = 0 0 0 1 The adjugate of a square matrx A s defned as the transpose of the matrx of cofactors: T adja = γ j It can be shown that AadjA= det A I ( ) adja If det A 0, we have A = I det A A square matrx A wth det A 0 s called nonsngular, for a nonsngular matrx we can defne the nverse adja = = = det A 1 1 1 A AA I, A A I

Rank Consder an m n matrx A. The number of lnearly ndependent rows of A equals the number of ts lnearly ndependent columns. The rank of A s the number of ts lnearly ndependent rows or columns. ( m n) rank A max, If A s a square matrx of sze n, rank A= n det A 0

MATLAB Basc Operatons + Addton A and B must have the same sze, unless one s a scalar. A scalar can be added to a matrx of any sze. - Subtracton A and B must have the same sze, unless one s a scalar. A scalar can be subtracted from a matrx of any sze. * Matrx multplcaton. For nonscalar A and B, the number of columns of A must equal the number of rows of B. A scalar can multply a matrx of any sze. / Slash or matrx rght dvson. B/A s roughly the same as B*nv(A). More precsely, B/A = (A'\B')'. \ Backslash or matrx left dvson. If A s a square matrx, A\B s roughly the same as nv(a)*b, except t s computed n a dfferent way.

MATLAB Basc Operatons ^ Matrx power. X^p s X to the power p, f p s a scalar. If p s an nteger, the power s computed by repeated squarng. If the nteger s negatve, X s nverted frst. ' Matrx transpose. A' s the lnear algebrac transpose of A. For complex matrces, ths s the complex conjugate transpose..' Array transpose. A.' s the array transpose of A. For complex matrces, ths does not nvolve conjugaton. Note: The slash\backslash operatons are the better than nv to solve lnear equatons.

MATLAB Basc Functons norm rank det trace nv matrx or vector norm matrx rank determnant sum of the dagonal elements matrx nverse

Applcatons of Matrces Matrces are mportant n many applcatons. One of the most mportant s the soluton of sets of smultaneous lnear equatons: a x + a x + a x = b 11 1 22 2 1n n 1 a x + a x + a x = b n1 1 n2 2 nn n 2 =,f det 0 = 1 Ax b A x A b Another applcaton s n the soluton of sets of smultaneous lnear ordnary dfferental equatons, for example, equatons lke () () () my + cy + ky = f t ρv + αy = g t can be put n the form x = Ax+ b t Lq + Cv + By = h t where A s a properly defned square matrx, and xb, are properly defned column vectors. ()

Smlarty Transformatons Sometmes t s useful to solve the equatons n a coordnate system that s dfferent from the orgnal problem formulaton. Any square nonsngular matrx T can be consdered a transformaton matrx. 1 x Tx, x T x () 1 1 () Lnear coordnate transformatons of column vectors are accomplshed va transformatons = = For example, under such a transformaton x = Ax+ b t x = T ATx + T b t Matrces transform under a change of 1 A T AT = Ths s called a smlarty transformaton. coordnates accordng to

Specal Matrces Smlarty transformatons are used to transform matrces nto a varety of specal forms (when possble). Among these are: a11 a 0 ann 0 1 0 0 0 0 1 lower companon: 0 0 0 0 0 1 an1 an2 a nn 0 22 dagonal:, upper trangular: a a a 0 a a 0 0 0 a 11 12 1n 22 2n nn

Advanced Topcs ~ Defer Egenvalues/Egenvectors Functons of Matrces Cayley-Hamlton Theorem Sngular Values

Summary Vectors Basc defntons & operatons lnear dependent\ndependent sets Matrces Defntons Algebrac operatons Determnants, Rank, Inverse Smlarty transformatons & specal matrc forms MATLAB functons