EL 625 Lecture 2. State equations of finite dimensional linear systems

Σχετικά έγγραφα
Partial Differential Equations in Biology The boundary element method. March 26, 2013

Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3

SCHOOL OF MATHEMATICAL SCIENCES G11LMA Linear Mathematics Examination Solutions

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Matrices and Determinants

ΗΜΥ 220: ΣΗΜΑΤΑ ΚΑΙ ΣΥΣΤΗΜΑΤΑ Ι Ακαδημαϊκό έτος Εαρινό Εξάμηνο Κατ οίκον εργασία αρ. 2

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Second Order Partial Differential Equations

Homework 8 Model Solution Section

Section 8.3 Trigonometric Equations

ECE 308 SIGNALS AND SYSTEMS FALL 2017 Answers to selected problems on prior years examinations

Example Sheet 3 Solutions

EE512: Error Control Coding

Finite Field Problems: Solutions

2 Composition. Invertible Mappings

Reminders: linear functions

Exercises 10. Find a fundamental matrix of the given system of equations. Also find the fundamental matrix Φ(t) satisfying Φ(0) = I. 1.

forms This gives Remark 1. How to remember the above formulas: Substituting these into the equation we obtain with

Capacitors - Capacitance, Charge and Potential Difference

CHAPTER 48 APPLICATIONS OF MATRICES AND DETERMINANTS

Space-Time Symmetries

Second Order RLC Filters

Inverse trigonometric functions & General Solution of Trigonometric Equations

D Alembert s Solution to the Wave Equation

6.1. Dirac Equation. Hamiltonian. Dirac Eq.

DESIGN OF MACHINERY SOLUTION MANUAL h in h 4 0.

Srednicki Chapter 55

the total number of electrons passing through the lamp.

Homework 3 Solutions

MATH423 String Theory Solutions 4. = 0 τ = f(s). (1) dτ ds = dxµ dτ f (s) (2) dτ 2 [f (s)] 2 + dxµ. dτ f (s) (3)

The Simply Typed Lambda Calculus

Forced Pendulum Numerical approach

DERIVATION OF MILES EQUATION FOR AN APPLIED FORCE Revision C

CRASH COURSE IN PRECALCULUS

Dr. D. Dinev, Department of Structural Mechanics, UACEG

6.003: Signals and Systems. Modulation

Parametrized Surfaces

Ηλεκτρονικοί Υπολογιστές IV

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

TMA4115 Matematikk 3

The kinetic and potential energies as T = 1 2. (m i η2 i k(η i+1 η i ) 2 ). (3) The Hooke s law F = Y ξ, (6) with a discrete analog

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

On the Galois Group of Linear Difference-Differential Equations

[1] P Q. Fig. 3.1

Higher Derivative Gravity Theories

Spherical Coordinates

1 String with massive end-points

Lifting Entry (continued)

ES440/ES911: CFD. Chapter 5. Solution of Linear Equation Systems

Differential equations

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

Statistical Inference I Locally most powerful tests

Potential Dividers. 46 minutes. 46 marks. Page 1 of 11

Concrete Mathematics Exercises from 30 September 2016

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

An Inventory of Continuous Distributions

Πανεπιστήµιο Κύπρου Πολυτεχνική Σχολή

( ) 2 and compare to M.

Congruence Classes of Invertible Matrices of Order 3 over F 2

ECE Spring Prof. David R. Jackson ECE Dept. Notes 2

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 19/5/2007

From the finite to the transfinite: Λµ-terms and streams

Problem Set 9 Solutions. θ + 1. θ 2 + cotθ ( ) sinθ e iφ is an eigenfunction of the ˆ L 2 operator. / θ 2. φ 2. sin 2 θ φ 2. ( ) = e iφ. = e iφ cosθ.

Lecture 34 Bootstrap confidence intervals

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Class 03 Systems modelling

( y) Partial Differential Equations

Areas and Lengths in Polar Coordinates

CHAPTER 25 SOLVING EQUATIONS BY ITERATIVE METHODS

Areas and Lengths in Polar Coordinates

Homomorphism in Intuitionistic Fuzzy Automata

EE101: Resonance in RLC circuits

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 6/5/2006

Approximation of distance between locations on earth given by latitude and longitude

2. THEORY OF EQUATIONS. PREVIOUS EAMCET Bits.

The Spiral of Theodorus, Numerical Analysis, and Special Functions

Numerical Analysis FMN011

Lecture 10 - Representation Theory III: Theory of Weights

ST5224: Advanced Statistical Theory II

Derivation of Optical-Bloch Equations

Πρόβλημα 1: Αναζήτηση Ελάχιστης/Μέγιστης Τιμής

4.6 Autoregressive Moving Average Model ARMA(1,1)

ADVANCED STRUCTURAL MECHANICS

k A = [k, k]( )[a 1, a 2 ] = [ka 1,ka 2 ] 4For the division of two intervals of confidence in R +

ω ω ω ω ω ω+2 ω ω+2 + ω ω ω ω+2 + ω ω+1 ω ω+2 2 ω ω ω ω ω ω ω ω+1 ω ω2 ω ω2 + ω ω ω2 + ω ω ω ω2 + ω ω+1 ω ω2 + ω ω+1 + ω ω ω ω2 + ω

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8 questions or comments to Dan Fetter 1

6.3 Forecasting ARMA processes

Dynamic types, Lambda calculus machines Section and Practice Problems Apr 21 22, 2016

Solution Series 9. i=1 x i and i=1 x i.

derivation of the Laplacian from rectangular to spherical coordinates

Solutions to Exercise Sheet 5

The challenges of non-stable predicates

Section 9.2 Polar Equations and Graphs

= {{D α, D α }, D α }. = [D α, 4iσ µ α α D α µ ] = 4iσ µ α α [Dα, D α ] µ.

Notes on the Open Economy

g-selberg integrals MV Conjecture An A 2 Selberg integral Summary Long Live the King Ole Warnaar Department of Mathematics Long Live the King

Main source: "Discrete-time systems and computer control" by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1

UNIVERSITY OF CALIFORNIA. EECS 150 Fall ) You are implementing an 4:1 Multiplexer that has the following specifications:

Transcript:

EL 625 Lecture 2 EL 625 Lecture 2 State equations of finite dimensional linear systems Continuous-time: ẋ(t) = A(t)x(t) + B(t)u(t) y(t) = C(t)x(t) + D(t)u(t) Discrete-time: x(t k+ ) = A(t k )x(t k ) + B(t k )u(t k ) y(t k ) = C(t k )x(t k ) + D(t k )u(t k ) state x(t) - vector of length n input u(t) - r output y(t) - m A,B,C and D are matrices of sizes n n, n r, m n and m r respectively If A,B,C,D are functions of time timevarying If these matrices are constant with time time-invariant

EL 625 Lecture 2 2 State Differential equations of circuits basic elements resistor, capacitor, inductor Resistor: v R (t) = R(t)i R (t) v R (t) : the voltage across the resistor i R (t) : the current through the resistor R(t) : the resistance A resistor is a zero-memory element i R R h h + v R Any circuit with only resistors (a purely resistive network) has zero-memory and is of zero order (needs no states to describe it)

EL 625 Lecture 2 3 Capacitor: i C (t) = q C (t), q C (t) = C(t)v C (t) where: i C (t) : the current through the capacitor q C (t) : the electrical charge in the capacitor v C (t) : the voltage across the capacitor C(t) C(t) dv C(t) dt : the capacitance = i C (t) dc(t) dt v C (t) If the capacitance does not change with time, C dv C(t) dt = i C (t) i C C h h + v C

EL 625 Lecture 2 4 Inductor: v L (t) = φ(t), φ(t) = L(t)i L (t) where: i L (t) : the current through the inductor v L (t) : the voltage across the inductor φ(t) : the flux stored in the inductor L(t) L(t) di L(t) dt : the inductance = v L (t) dl(t) dt i L (t) If the inductance does not change with time, L di L(t) dt = v L (t) i L L h h + v L

EL 625 Lecture 2 5 Example: '$ + &% e (t) + v L (t) v 2 (t) + L i L (t) + v (t) C R i (t) C 2 R 2 i 2 (t) '$ + &% e 2 (t) Applying Kirchoff s current and voltage laws, i L = i + i 2 e = v L + i R + v e = v L + v 2 + i 2 R 2 + e 2 From the terminal relationships of the capacitors and the inductor, v = C i v 2 = C 2 i 2 i L = L v L

EL 625 Lecture 2 6 The inputs are e and e 2 The output is the voltage across the inductor, v L Choosing as our states, i L, v and v 2, x = i L v v 2, u = e e 2, y = [v L ] Need to express ẋ in terms of x and the inputs, e and e 2, i L = L v L But, v L is not a state variable we need to express v L in terms of the state variables and the inputs v L = i R R 2 L +v R 2 +v R 2 +e +e R 2 R + R 2 R + R 2 R + R 2 R + R 2 i L R R 2 R 2 R = i L + v + v 2 L(R + R 2 ) L(R + R 2 ) L(R + R 2 ) +e R + e 2 L L(R + R 2 )

EL 625 Lecture 2 7 Similarly, i = i L R 2 R + R 2 + v R + R 2 + v 2 R + R 2 + e 2 R + R 2 v = C i R 2 = i L + v + v 2 C (R + R 2 ) C (R + R 2 ) C (R + R 2 ) +e 2 C (R + R 2 ) i 2 = i L R R + R 2 + v R + R 2 + v 2 R + R 2 + e 2 R + R 2 v 2 = C 2 i 2 R = i L + v C 2 (R + R 2 ) +e 2 C 2 (R + R 2 ) C 2 (R + R 2 ) + v 2 C 2 (R + R 2 )

EL 625 Lecture 2 8 ẋ = Ax + Bu y = Cx + Du where A = B = R R 2 L(R +R 2 ) R 2 C (R +R 2 ) R C 2 (R +R 2 ) L R L(R +R 2 ) C (R +R 2 ) C 2 (R +R 2 ) R 2 L(R +R 2 ) C (R +R 2 ) C 2 (R +R 2 ) R L(R +R 2 ) C (R +R 2 ) C 2 (R +R 2 ) C = [ R R 2 R 2 R R +R 2 R +R 2 R +R 2 ] D = [ R R +R 2 ]

EL 625 Lecture 2 9 State Differential Equations of Mechanical Systems Spring: f K (t) = K(t)[z 2 (t) z (t)] Hooke s Law where: z (t) and z 2 (t) : the displacements of the two ends of the spring f K (t) : the force applied K(t) : the spring constant K z z 2 f K Damping Element: f D (t) = D(t)[v 2 (t) v (t)] where: v (t) and v 2 (t) : the velocities of the two ends of the damping element f D (t) D(t) : the force applied : the damping coefficient v v 2 f D D

EL 625 Lecture 2 Mass: M(t) dv(t) dt = f M (t) dm(t) dt v(t) where: v(t) : the velocity of the mass f M (t) : the force applied M(t) : the mass If mass does not change with time, M dv(t) dt = f M (t) M v f M

EL 625 Lecture 2 Example: D M x x 2 K M 2 f(t) f(t) is the input and x 2 (t) is the output u = [f(t)],y = [x 2 ] Choose as the state variables x, x 2, v and v 2 where v = ẋ and v 2 = ẋ 2 = x = x x 2 v v 2 M 2 v 2 = f(t) + K(x x 2 ) M v = K(x 2 x ) + D( v )

EL 625 Lecture 2 2 ẋ = Ax + Bu y = Cx + Du where A = B = C = [ K M K M 2 M 2 K M D M K M 2 ] D = []

EL 625 Lecture 2 3 Choice of state variables is not unique Let x a (t) be a valid set of state variables with ẋ a = A a x a + B a u y = C a x a + D a u () Consider x b (t) = T x a (t) where T is an n n nonsingular matrix ẋ b = T ẋ a = T A a T x b + T B a u y = C a T x b + D a u (2) x b (t) is also a valid set of state variables! where: ẋ b = A b ẋ b + B b u y = C b x b + D b u (3) A b = T A a T B b C b = T B b = C a T D b = D a (4) Similarity Transformation: x b = T x a

EL 625 Lecture 2 4 Convenient choice of state variables : inductor currents and capacitor voltages for fixed networks inductor fluxes and capacitor charges for time-varying networks differences in displacements of the ends of springs from their equilibrium positions and velocities of masses for mechanical systems Simulation Diagrams: Basic Simulator Elements: Dynamic element (a) integrator : for analog systems t u(t) y(t) = y(t ) + t u(τ) dτ y = D u (b) delay : for discrete-time systems u(t k ) y(t k ) = u(t k ) y = E u

EL 625 Lecture 2 5 2 Summing element - adder u (t) u 2 (t) u r (t) + + + '$ &% y(t) = r u i (t) i= 3 Scaling Element (Amplifier or Attenuator) u(t) K(t) y(t) = K(t) u(t) Minimal Realization: Fewest possible number of dynamic elements Convenient choice of state variables Outputs of integrators and delay elements Example: y + 3tÿ + 2ẏ + α(t)y = ü + e t u + u D 3 y + 3tD 2 y + 2Dy + α(t)y = D 2 u + e t Du + u (D is the derivative operator) y = D 3 { D 2 u 3tD 2 y + e t Du 2Dy + u α(t)y } y = D { D 2 (D 2 u 3tD 2 y) + D { D (e t Du 2Dy)

EL 625 Lecture 2 6 +D {u α(t)y} }} = D { u D 2 (3tD 2 y) + D { D (e t Du) 2y +D {u α(t)y} }} Using integration by parts, D 2 (3tD 2 y) = D (3tDy D (3Dy)) = D (3tDy 3y) = 3ty D (3y) D (3y) = 3ty D (6y) D (e t Du) = e t u + D (ue t ) Thus, y = D { u 3ty + D 6y + D { e t u + D (ue t ) 2y +D {u α(t)y} }} = D { u 3ty + D { e t u + +4y + D {u + ue t α(t)y} }}

EL 625 Lecture 2 7 Simulation Diagram: u + e t α(t) 4 3t + + + + x 3 x + 2 + x y e t Choosing the outputs of the integrators as the states, we have ẋ = u 3tx + x 2 ẋ 2 = 4x + x 3 + e t u A(t) = x = 3t 4 α(t) x x 2 x 3 ẋ 3 = u( + e t ) α(t)x y = x ; B(t) = e t + e t ẋ = A(t)x + B(t)u ; C(t) = y = C(t)x + D(t)u T ; D(t) = []

EL 625 Lecture 2 8 Example: y(k + 3) + 3ky(k + 2) + 2y(k + ) + α(k)y(k) = u(k + 2) + e k u(k + ) +u(k) E 3 y(k) + 3kE 2 y(k) + 2Ey(k) + α(k)y(k) = E 2 u(k) + e k Eu(k) +u(k) where E is the delay operator y(k) = E 3{ E 2 u(k) 3kE 2 y(k) + e k Eu(k) 2Ey(k) + u(k) α(k)y(k) } = E { E 2 (E 2 u(k) 3kE 2 y(k)) + E { E (e k Eu(k) 2Ey(k)) +E {u(k) α(k)y(k)} }} E 2 (3kE 2 y(k)) = 3(k 2)y(k) E (e k Eu(k)) = e k u(k) Thus, y(k) = E { u(k) 3(k 2)y(k) + E { e (k ) u(k) 2y(k) +E {u(k) α(k)y(k)} }}

EL 625 Lecture 2 9 Simulation Diagram: u(k) α(k) 2 3(k 2) + + + x 3 x + 2 + x y(k) e (k ) Choosing the outputs of the delay elements as the states, we have x (k + ) = u(k) 3(k 2)x (k) + x 2 (k) x 2 (k + ) = e (k ) u(k) 2x (k) + x 3 (k) x 3 (k + ) = u(k) α(k)x (k) y(k) = x (k) x(k) = x (k) x 2 (k) x 3 (k) x(k + ) = A(k)x(k) + B(k)u(k) y(k) = C(k)x(k) + D(k)u(k)

EL 625 Lecture 2 2 A(k) = 3(k 2) 2 α(k) B(k) = e (k ) C(k) = [ ] D(k) = []

EL 625 Lecture 2 2 Simpler method if no derivatives of the input are in the equation: y (n) +α n (t)y (n ) +α n 2 (t)y (n 2) + +α (t)y () +α (t)y = β(t)u(t) (y (i) i th derivative of y(t)) Choose, x = y x 2 = y () x 3 = y (2) x n = y (n ) x = x x 2 x 3 x n ẋ = x 2 ẋ 2 = x 3 ẋ n = x n ẋ n = α n (t)x n α n 2 (t)x n α (t)x y = x +β(t)u(t)

EL 625 Lecture 2 22 ẋ = α α α n 2 α n } {{ } A A is in companion matrix form y = [ ] } {{ } C x + x + [] u }{{} D β } {{ } B Another method: Let α, α,, α n and β be constants D n y + α n D n y + α n 2 D n 2 y + + α Dy + α y = βu u D n y = α n D n y α n 2 D n 2 y α Dy α y + βu y = D n { α n D n y α n 2 D n 2 y α Dy α y + βu } = D { α n y + D { α n 2 y + D { +D ( α y + βu ) } }} }{{} n

EL 625 Lecture 2 23 ż = α n z + z 2 ż 2 = α n 2 z + z 3 z = ż n ż n = α z + x n = α z + βu z z 2 z n z n y = z z = α n α n 2 α 2 α α } {{ } A [ y = z + β } {{ } B z + [] }{{} C ] }{{} u D u This method gave different A and B matrices - z and x are related through a similarity transformation

EL 625 Lecture 2 24 where T = z = T x α n α 2 α n α α n It can be checked that A = T AT, B = T B, C = CT and D = D The D matrix does not change under a similarity transformation A non-singular D matrix = the impulse response has an impulse

EL 625 Lecture 2 25 MIMO systems: ÿ + t 2 ẏ 2 + ẏ + ty + y 2 = tu + u 2 ÿ 2 + tẏ + y 2 y = t u + u 2 From the first equation, D 2 y + t 2 Dy 2 + Dy + ty + y 2 = tu + Du 2 y = D 2{ t 2 Dy 2 Dy + Du 2 + tu ty y 2 } = D { D ( t 2 Dy 2 Dy + Du 2 ) + D {tu ty y 2 } } D (t 2 Dy 2 ) = t 2 y 2 D (2ty 2 ) Thus, y = D { u 2 y t 2 y 2 + D {2ty 2 + tu ty y 2 } } Similarly, from second equation, D 2 y 2 + tdy + y 2 y = tdu + u 2 y 2 = D 2{ tdu tdy + u 2 + y y 2 } = D { D (tdu tdy ) + D {u 2 + y y 2 } }

EL 625 Lecture 2 26 D (tdu ) = tu D u D (tdy ) = ty D y y 2 = D { tu ty + D {2y y 2 u + u 2 } } State-space realization: ẋ = x +x 2 t 2 x 3 +u 2 ẋ 2 = tx +(2t )x 3 +tu ẋ 3 = tx +x 4 +tu ẋ 4 = 2x x 3 u +u 2 y = x y 2 = x 2

EL 625 Lecture 2 27 Simulation Diagram u 2 u t t + + + x 2 + x y 2t t 2 + 2 + t x 4 + + x 3 y 2 t