CDS 101: Lecture 8.2 Tools for PID & Loop Shaping

Similar documents
EE Control Systems

Course Outline. Designing Control Systems. Proportional Controller. Amme 3500 : System Dynamics and Control. Root Locus. Dr. Stefan B.

Lab(8) controller design using root locus

Mechatronics. Time Response & Frequency Response 2 nd -Order Dynamic System 2-Pole, Low-Pass, Active Filter

CDS 101/110a: Lecture 8-1 Frequency Domain Design

CDS 101: Lecture 5.1 Controllability and State Space Feedback

Chapter 2 Feedback Control Theory Continued

Answer: 1(A); 2(C); 3(A); 4(D); 5(B); 6(A); 7(C); 8(C); 9(A); 10(A); 11(A); 12(C); 13(C)

CHAPTER NINE. Frequency Response Methods

MEM 255 Introduction to Control Systems: Analyzing Dynamic Response

2.004 Dynamics and Control II Spring 2008

Bode Diagrams School of Mechanical Engineering ME375 Frequency Response - 29 Purdue University Example Ex:

732 Appendix E: Previous EEE480 Exams. Rules: One sheet permitted, calculators permitted. GWC 352,

Lecture 13. Graphical representation of the frequency response. Luca Ferrarini - Basic Automatic Control 1

Crash course part 2. Frequency compensation

1the 1it is said to be overdamped. When 1, the roots of

Frequency Response Methods

School of Mechanical Engineering Purdue University. ME375 Frequency Response - 1

Analysis of the No-Load Characteristic of the Moving Coil Linear Compressor

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Mechanical Engineering Dynamics and Control II Fall K(s +1)(s +2) G(s) =.

Dynamic Response of Linear Systems

EE 4343 Lab#4 PID Control Design of Rigid Bodies

ME 375 FINAL EXAM Friday, May 6, 2005

Dr. Seeler Department of Mechanical Engineering Fall 2009 Lafayette College ME 479: Control Systems and Mechatronics Design and Analysis

Chapter 7: The z-transform. Chih-Wei Liu

CONTROL SYSTEMS. Chapter 7 : Bode Plot. 40dB/dec 1.0. db/dec so resultant slope will be 20 db/dec and this is due to the factor s

Voltage controlled oscillator (VCO)

Class 07 Time domain analysis Part II 2 nd order systems

Course Outline. Problem Identification. Engineering as Design. Amme 3500 : System Dynamics and Control. System Response. Dr. Stefan B.

CDS 101/110a: Lecture 10-1 Robust Performance

Run-length & Entropy Coding. Redundancy Removal. Sampling. Quantization. Perform inverse operations at the receiver EEE

Contents Kreatryx. All Rights Reserved.

Introduction to Control Systems

ECE 422/522 Power System Operations & Planning/Power Systems Analysis II : 6 - Small Signal Stability

Automatic Control Systems

The z-transform. 7.1 Introduction. 7.2 The z-transform Derivation of the z-transform: x[n] = z n LTI system, h[n] z = re j

EE 505. Lecture 28. ADC Design SAR

FIR Filter Design: Part II

Introduction to Signals and Systems, Part V: Lecture Summary

6.003 Homework #3 Solutions

Sinusoidal Steady-state Analysis

21. Control Systems. Easy: Incremental Linear: P PI PID Intuitive: Fuzzy Logic. It says to cook your food until burned food!

Control Systems I Lecture 10: System Specifications

THE SOLUTION OF NONLINEAR EQUATIONS f( x ) = 0.

Outline. Classical Control. Lecture 1

EE C128 / ME C134 Fall 2014 HW 6.2 Solutions. HW 6.2 Solutions

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Mechanical Engineering Dynamics and Control II Fall 2007

Exponential Moving Average Pieter P

Butterworth LC Filter Designer

Chapter 2. Classical Control System Design. Dutch Institute of Systems and Control

Maximum and Minimum Values

(s)h(s) = K( s + 8 ) = 5 and one finite zero is located at z 1

Methods for analysis and control of. Lecture 4: The root locus design method

Chapter 4 : Laplace Transform

CDS 101/110a: Lecture 10-2 Control Systems Implementation

UNIVERSITY OF BOLTON SCHOOL OF ENGINEERING BENG (HONS) IN MECHANICAL ENGINEERING SEMESTER 1 EXAMINATION 2016/2017

Ch3 Discrete Time Fourier Transform

ECEN620: Network Theory Broadband Circuit Design Fall 2014

Exercises for lectures 13 Design using frequency methods

Spectral Analysis. This week in lab. Next classes: 3/26 and 3/28. Your next experiment Homework is to prepare

System Diagnostics using Kalman Filter Estimation Error

Easy Tuning of PID Controllers for Specified Performance

Frequency methods for the analysis of feedback systems. Lecture 6. Loop analysis of feedback systems. Nyquist approach to study stability

Chapter 6 - Solved Problems

CDS 101: Lecture 4.1 Linear Systems

Systems Analysis and Control

MA Lesson 26 Notes Graphs of Rational Functions (Asymptotes) Limits at infinity

3. Z Transform. Recall that the Fourier transform (FT) of a DT signal xn [ ] is ( ) [ ] = In order for the FT to exist in the finite magnitude sense,

Lecture 5: Frequency domain analysis: Nyquist, Bode Diagrams, second order systems, system types

Engraving Machine Example

Dynamic System Response

Transfer func+ons, block diagram algebra, and Bode plots. by Ania- Ariadna Bae+ca CDS Caltech 11/05/15

PID SELF-TUNING CONTROLLERS USED IN DIFFERENT STRUCTURES OF THE CONTROL LOOPS

Control Systems I. Lecture 9: The Nyquist condition

CDS 101/110: Lecture 3.1 Linear Systems

CALCULUS BASIC SUMMER REVIEW

Lecture 3. Digital Signal Processing. Chapter 3. z-transforms. Mikael Swartling Nedelko Grbic Bengt Mandersson. rev. 2016

Transient response via gain adjustment. Consider a unity feedback system, where G(s) = 2. The closed loop transfer function is. s 2 + 2ζωs + ω 2 n

A New Accurate Analytical Expression for Rise Time Intended for Mechatronics Systems Performance Evaluation and Validation

Systems Analysis and Control

Computational Fluid Dynamics. Lecture 3

Chapter 7 z-transform

OBJECTIVES. Chapter 1 INTRODUCTION TO INSTRUMENTATION FUNCTION AND ADVANTAGES INTRODUCTION. At the end of this chapter, students should be able to:

100 (s + 10) (s + 100) e 0.5s. s 100 (s + 10) (s + 100). G(s) =

Lyapunov Stability Analysis for Feedback Control Design

Jitter Transfer Functions For The Reference Clock Jitter In A Serial Link: Theory And Applications

Mechatronics II Laboratory Exercise 5 Second Order Response

AMME3500: System Dynamics & Control

Systems Analysis and Control

5. Fast NLMS-OCF Algorithm

Systems Analysis and Control

APPLICATIONS FOR ROBOTICS

Dynamic Response of Second Order Mechanical Systems with Viscous Dissipation forces

Automatic Control 2. Loop shaping. Prof. Alberto Bemporad. University of Trento. Academic year

Course roadmap. ME451: Control Systems. What is Root Locus? (Review) Characteristic equation & root locus. Lecture 18 Root locus: Sketch of proofs

FIR Filter Design: Part I

Chapter 9: Numerical Differentiation

Exam. Notes: A single A4 sheet of paper (double sided; hand-written or computer typed)

Methods for analysis and control of dynamical systems Lecture 4: The root locus design method

EE 435. Lecture 25. Data Converters. Architectures. Characterization

Transcription:

CDS : Lecture 8. Tools for PID & Loop Shapig Richard M. Murray 7 November 4 Goals: Show how to use loop shapig to achieve a performace specificatio Itroduce ew tools for loop shapig desig: Ziegler-Nichols, root locus, lead compesatio Work through some eample cotrol desig problems Readig: Åström ad Murray, Aalysis ad Desig of Feedback Systems, Ch 8

Tools for Desigig PID cotrollers d r - e C(s) u P(s) y Cs () = K( Ts D ) Ts I Zeigler-Nichols tuig Desig PID gais based o step respose Works OK for may plats (but uderdamped) Good way to get a first cut cotroller Frequecy domai versio also eists Cautio: PID amplifies high frequecy oise Sol : pole at high frequecy Cautio: Itegrator widup Prologed error causes large itegrated error Effect: large udershoot (to reset itegrator) Sol : move pole at zero to very small value Facier sol : ati-widup compesatio K =. / a T = * L T = L/ 7 Nov 4 R. M. Murray, Caltech CDS a Magitude (db) Step respose 8 6 4 L I Poit of maimum slope Bode Diagrams -4-4 Frequecy (rad/sec) D

Eample: PID cruise cotrol Ps () / m r = s b/ m s a.5.4.3 L =.49 a = step.39 Ziegler-Nichols desig for cruise cotroller Plot step respose, etract L ad a, compute gais Bode Diagrams.. slope -. 3 4 5 Phase (deg); Magitude (db) 5-5 - - Ps () Ps () Cs () - -3 - - Frequecy (rad/sec) Cs () L() s L() s Time (sec.) Result: sluggish icrease loop gai 7 Nov 4 R. M. Murray, Caltech CDS 3 Amplitude.5.5 Step Respose Cs () = K( Ts D ) Ts K =. / a T = * L T = L/ I 3 4 5 6 I D

r - e Pole Zero Diagrams ad Root Locus Plots C(s) α Pole zero diagram verifies stability Roots of PC give closed loop poles Ca trace the poles as a parameter is chaged: Cs () = K( Ts D ) Ts I α u P(s) y Imag Ais -8-7 -6-5 -4-3 - - 3 Real Ais Root locus = locus of roots as parameter value is chaged Ca plot pole locatio versus ay parameter; just repeatedly solve for roots Commo choice i cotrol is to vary the loop gai (K) 8 6 4 - -4-6 Origial pole locatio (α = ) Pole goes to Poles merge ad split Pole goes ustable for some α Pole goes to termial value 7 Nov 4 R. M. Murray, Caltech CDS 4

Oe Parameter Root Locus Basic idea: covert to stadard problem : as () αbs () = Look at locatio of roots as α is varied over positive real umbers If phase of a(s)/b(s) = 8, we ca always choose a real α to solve eq Ca compute the phase from the pole/zero diagram Gs () a() s ( s z)( s z) L( s zm) = = k b() s ( s p )( s p ) L( s p ) G( s ) = ( s z ) L ( s z ) ( s p ) L ( s p ) m φ i = phase cotributio from s to -p i ψ I = phase cotributio from s to -z i Trace out positios i plae where phase = 8 At each of these poits, there eists gai α to satisfy a(s) αb(s) = All such poits are o root locus 7 Nov 4 R. M. Murray, Caltech CDS 5

r - e α C(s) Loop gai as root locus parameter Commo choice for cotrol desig Special properties for loop gai Roots go from poles of PC to zeros of PC Ecess poles go to ifiity Ca compute asymptotes, break poits, etc Very useful tool for cotrol desig MATLAB: rlocus Root Locus for Loop Gai u P(s) s () α d( s) α( s) = d() s y Imag Ais 8 6 4 - -4-6 Ope loop pole locatio (α = ) Asymptotes for ecess poles at (36 /(P-Z)) Real ais to the left of odd # of real poles & zeros is o root locus Closed pole goes to ope loop zeros -8-7 -6-5 -4-3 - - 3 Real Ais Additioal commets Although loop gai is the most commo parameter, do t forget that you ca plot roots versus ay parameter Need to lik root locatio to performace 7 Nov 4 R. M. Murray, Caltech CDS 6

Secod Order System Respose Secod order system respose Sprig mass dyamics, writte i caoical form Guidelies for pole placemet Dampig ratio gives Re/Im ratio Settig time determied by Re(λ) H() s ω ω = = s ςω s ω ( s ςω jω )( s ςω jω ) d d ω = ω ς d Performace specificatios T T r s.8/ ω 3.9 / ςω M e p e = SS πς / ς 8 6 4 Root Locus Editor (C) T s < ζ.77.5 M p 4% 6% Slope - -.7 Imag Ais - -4 Desired regio for closed loop poles M p < y.5 44% -3.9-6 -8 - - -8-6 -4 - - -8-6 -4-7 Nov 4 R. M. Murray, Caltech CDS Real Ais 7

Effect of pole locatio o performace Idea: look at domiat poles Poles earest the imagiary ais (earest to istability) Aalyze usig aalogy to secod order system PZmap complemets iformatio o Bode/Nyquist plots Similar to gai ad phase calculatios Shows performace i terms of the closed loop poles Particularly useful for choosig system gai Also useful for decidig where to put cotroller poles ad zeros (with practice [ad SISOtool]) Imag Ais Imag Ais Imag Ais.5.5 -.5 - -.5-5 -4-3 - - Real Ais.5.5 -.5-3 4 5 6 7 8 9 Time (sec.) 7 Nov 4 R. M. Murray, Caltech CDS 8 Pole-zero map -.5-5 -4-3 - - Real Ais.5.5 -.5 - Pole-zero map Pole-zero map -.5-5 -4-3 - - Real Ais Amplitude Amplitude Amplitude To: Y() To: Y() To: Y().4..8.6.4..4..8.6.4. 3.5.5.5.5 Step Respose From: U() Step Respose From: U() 3 4 5 6 7 8 9 Time (sec.) 3 Step Respose From: U() 3 4 5 6 7 8 9 Time (sec.)

Eample: PID cruise cotrol Start with PID cotrol desig: / m r Ps () = s b/ m s a Cs () = K( Ts D ) Ts I Modify gai to improve performace Use MATLAB sisotool Adjust loop gai (K) to reduce overshoot ad decrease settlig time ζ less tha 5% overshoot Re(p) < -.5 T s less tha sec 7 Nov 4 R. M. Murray, Caltech CDS 9

Eample: Pitch Cotrol for Caltech Ducted Fa System descriptio Vector thrust egie attached to wig Iputs: fa thrust, thrust agle (vectored) Outputs: positio ad orietatio States:, y, θ derivatives Dyamics: flight aerodyamics Cotrol approach Desig ier loop cotrol law to regulate pitch (θ ) usig thrust vectorig Secod outer loop cotroller regulates the positio ad altitude by commadig the pitch ad thrust Basically the same approach as aircraft cotrol laws 7 Nov 4 R. M. Murray, Caltech CDS

Phase (deg); Magitude (db) 5 Performace Specificatio ad Desig Approach Performace Specificatio % steady state error Zero frequecy gai > % trackig error up to rad/sec Gai > from - rad/sec 45 phase margi Gives good relative stability Provides robustess to Frequecy (rad/sec) ucertaity r Desig approach Ps () = Js ds mgl Ope loop plat has poor phase margi Add phase lead i 5-5 rad/sec rage a = 5 Icrease the gai to achieve steady s a Cs () = K b = 3 state ad trackig performace specs s b K = 5 3 Avoid itegrator to miimize phase -5-5 - -5-7 Nov 4 R. M. Murray, Caltech CDS

Summary: PID ad Root Locus PID cotrol desig Very commo (ad classical) cotrol techique Good tools for choosig gais u = K e K e K e& p I D Bode Diagrams Root locus Show closed loop poles as fuctio of a free parameter Performace limits RHP poles ad zeros place limits o achievable performace Waterbed effect Phase (deg); Magitude (db) 5 4 3 5 ω = T I ω = T D Imag Ais 8 6 4 - -4-5 -6 - -3 - - 3 Frequecy (rad/sec) -8-7 -6-5 -4-3 - - 3 Real Ais 7 Nov 4 R. M. Murray, Caltech CDS