Beruflich Dokumente
Kultur Dokumente
lv
tv
tu
E
F A Project Report
E
t
Positional Control System- Analysis &
Design
frojtareyert"tufitnitte{inyartialfttffi ttnentof tfrereyuirement
lj
k fertde ora"{"frf,eg"ee
"f
E Bachelor of Technology in Electrical
2011
Engineering
t
l_.
k
By
Dehojit tvlaitra
Sai"fr.atg fiatt a.cfiarrya
E
L.
lj
tuati.fr,Dey
' Sen"dniSioSf*
Defiottarn Cfrnfrrafiorty
k SryiiN Sfi"umc
E
k Under the Guidance of
k ejry Cfrmda
l- "rof.
E
l_.
Department of Electrical Engineering
JIS College of Engineering
E
Block-A, Phase-III, Kalyani, Nadia
747295
l-t
L-
l-.
l-
t:
l:
l:
l:
L-
ll ,
CERTIflGJTTE
Prof.A. Chanda
(ProjectGuide)
AssistantProfessor
Dept.Of ElectricalEngineering
Kalyani.
JISCollegeof Engineering,
S
1--
lirr
t\-
h-,
-
- J15College
of Engineering
8lo* 'A', PhaseJll.Kalyani,H:dia,741335
- Fhone:+91 33 258? 2137,Teh{sx: +91 33 ?582:138
Websir*:www-jiscalt+je-ae.in,Fmail: in@;iisnollege-m.in
-
-
-
- OF*.PPROV*.L
CENTIflC^ETE,
:-
t Thisis to certifythat the projectentitled"POSITIONAI
CONTROL
SYSTEM-
ANALYSIS
&
t
t
submittedby DebojitMaitra,SaikatBhattachotya,PratikDey,SandhiSingha,
DESfGN"
t DebottamChakraborty,
SaniivShomeof Electrical
Engineering
Department,
JlSCollege
t
t
Kalyani,in the partialfulfillmentof the requirementfor the awardof
of Engineering,
, the degree of Bachelorof ElectricalEngineeringin West Bengal Universityof
- -
(WBUT)duringthe academic
Technology year2AtA-20Lt.
D
t
=
=
-:
a,
(Prof. Asit,5*6x Guha) (Prof. U. Banerjee)
t
t Directo Principal
t JIS Collegeof En JIS College of Engineering
Kalvani - 74 Kalyani - 741235
-
t
t
o
o
o (Prof. Subrata K (Prof. A Chanda)
?
Head Guide, Lecturer
-t
El ectrical Engineering Departuent Electrical Engineering Department
{
JIS College of Engineering JIS College af Engineering
-t
Kalyani - 741235 Kalvani - 741235
t
a
-t
t Corpotere offirel 7, larat 8sr€ Eoad. Iclletn-IlHl SlU, Plroneq+91 33 ll89 394#5323, Tdefexl +91 l$ 22*9 3945
{
rt
ra-
t
D
t
-
I
"*.glilOIT/I.uDqEITUNT
We would like to extend our sincere gratitude to our respected teacher Mr" A Chanda for his
assistanceand guidance towards the progress of this project.
t We also appreciate with deep senseof thankfulness, the co- operation extended to us by the lab
t
assistantand some of our frierrds and other teachers.
t
-
Lastly we acknowledge our indebtednessto the principal of institution and our parents whose
support, dedication have given us an immense help in commencing this project.
We have made a humble attempt in this direction to make this project perfect as far as possible.
In spite of efforts some errors might.crept in. we shall be grateful to the teacher if some are
brought by notice. Comments and criticism is welcome in this regard.
Dat e:
t elo s l u t
Place:f",Iry *;
Debojit Maitra
)"u,b,#t l.t^[c" .
(07r230n601s)
Saikat Bhattacharya
3 d,raf B\^GU64-Lorla ,
(07r230n6016)
Pratik Dey
(07r230n6017)
Sandhi Singha
(07r230n6018)
Debottam Chakraborty
(07r230n6019)
Sanjiv Shome !r
Sorn;tv ShPcrno,
(07r230n6020)
f,
J
GONTENT
{
Title page
{ Certificate I
H Certificate of approval II
Acknowledgement III
-l
it Content ry
V
I 1. fntroduction
I 1
1"1 Introduction
? L.2 2
Objectiveof this project
:t 1.3 3
Basic conceptof positional control
4
1.8.1 Detectionof positioning errors during movement
I.4 4
MATLAB
4
r"4.L category: Generalpu{poseapplicationsand utilities
available on-all systems
4
L.4.2 Brief descriptionof iVIATLAB prod.ucts
5
1.4.3 SIMULINK.
o
I.4.4 What is a SIMULINK function
6
2. Proportional integral-derivative control
& lead -lag compensator
2.1 Introduction 6
7
2.2 Different types of controller
7
2.2.L Proportionalcontrol(Gain)
7
2.2.2 Integral control (Automatic reset)
8
2.2.3 Derivative control (pre Act(TM) or rate)
10
2.2.4 Putting it together:pID Controller
T2
Lead-LagCompensator
13
2.3.L Application
2.3.2 13
Theory
2.3.3 13
Implementation
2.3.4 L4
IntuitiveExplanation
15
2.4 Time responsespecifications
15
2.4.t Overshoot(Signal)
15
2.4.L.I Definition
16
2.4.1.2 Controltheory
16
fr
V l_
t
2.4.2 Rise time 16
2.4.2.1 Application 18
2.4.3 Settling time 1_9
2.4.3.L Application 2A
2"4.4 Steadystate error 2t
2.4.4.1 Application 2L
2.5 Importance of PID and lead-lagcompensatorover
time responsespecifications 22
2.5.L Introduction 22
2.5.2 Three -term controller 23
2.5.3 The characteristicsof p, i, and d controllers 24
4.2 Robotics 43
4.2.1 Introduction 43
4.2.2 Experiment 44
tt
lvl
t,
4.3 Radar control 45
4.3.1 Introduction 45
4"3.2 The systemapproach 45
Conclusion 46
References 48
fr
vll
t,
INTRODUCTION
E
l-
l'r Introduction
It Chapter1
l-
It
!-1
It
f-1
I.1 INTRODUCTION
It
F' This project confines itself to a study of dynamic and steady
state modeling of seryo
fF! system,henceprogrammingof the dc motor and analysis
of its different dynamic parameters
like settlingtimg rise time, peakovershoot,steadystateeror
F:) by using MATLAB tools.
r!
t The dynamic simulation procedureis derived considering
a state of data of DC drive
b
t\
from the subsystemdifferential and functional relationships"
Attempt has been made to
D
\ undertakethe problan of designingin MATLAB Programing
a and 5IMULINK e,nvironment
for a closed loop control system. In order to put the idea
? on a familiar footing we have
? considerthe designof PID,PI & PD controllerby writing MATLAB
script then simulatethe
b dynamicsusing SIMULINK library functions blocks. A position
conhol system employing
D an armatureconfrolledDC servomotorand a DC drive related
to designcriteria towardstotal
i' nanative designprocedureof the systernhas been undertaken.
T As in many casesof system
T design,there is no unique solution to the design problem
of a pID controller. However
;' attempthasbeenmadetowardsthe objectivein the contoller
designis to selectthe suitable
t valuesfor the parameterskp, ki & kd ts satisff the desiredrequirements"
t
t
D
)
t
t
I
ir
)
)
)
)
t,
z
\
l Introduction Chapter1
I
L.2 OBJECTI1M
t
3
t
h-
-t
hf
D
Introduction Chapter1
-:
r{
e 1.3 BASIC CONCEPT OF POSITIONAL CONTROL
\
D
A positional control is a type of automaticconhol in which the input commandsare the
!
-t desiredpositionof a body. If a body or a systemmovesfrom its desiredposition,it makesan
= error. Through this control system,we can eliminate this positional error. By the use of pID
= conholler and leadJag compensatorwe can determinethe exact value of this error and
.!
t
optimizethe system.
D
D
D
: 1.3.1.DETECTION OF POSITIONING ERRORS DT]RING
:
? MO\MMENTS
? Maximumtolerableerror may vary (canbe as small as25 pm)
?
? Drive speedlimit = max error/ systemresponsetime
I E.g.25 pm / 100ms = 0.25mm/s
?
{ In casethis is in the resonancgrange of the assernbly,the speedshould be further
r-
reduced.(Continuoussinglestepmode).
r!
t
,
t 1.4 MATI"AB
i{
t
q,
D MATLAB is an integratedtechnical computing environmentthat combines numeric
D computation,
advancedgraphicsandvisualization,andahigh-levelprograrnminglanguage.
D
D
D
1.4.L CATEGORY:GENERAL PIIRPOSE APPLICATIONS
,
D AIYD UTILITTES AVAII,ABI,E ON-.ALL SYSTEMS
:
? Short for Matrix Laboratory MATLAB is a technical computing environment for
? high-performancenumeric computationand visualization. It integratesnumerical analysis,
t mahix computation,signal processing and graphics in an easy-to-useenvironment where
{
.a problems and solutions are expressedjust as they are written mathematically, without
q haditional programming. Tlpical uses include general-purposenumeric computation,
,
q
t algorithm prototyping, and special-purposeproblem solving with matrix formulations that
l\
,
D
t
D t
4 t____ _ _'.__--
) 'a
t
,
i-
I
lr-
*. Introduction Chapter1
-
:t
rr{
arise in disciplinessuch as linear algebra,structuralanalysis,statistics,and digital signal
rrra
t processing.
- MATLAB is included in the default user path. The current version can be evoked by
-
tping'MATLAB" on the commandline.
-
\i
- -
,
1.4.2 BRIEF DESCRIPTION OF IT{ATI"AB PRODUCTS
, . The following tools arein MATLAB, which makesthe softwaremore easyanduseful for
, anykind of MathematicalandProgrammingworks.TheMATLB tools are----
J
1. CDMA ReferenceBlockset
-
a 2. Communications
Toolbox
\ >
J
3. ControlSystemToolbox
-
a
4. D ataAcquisitionToolbox
-t 5. Database
Toolbox
't 6. Developer's
Kit for TexasInstruments
DSP
-:
J 7. InstrumentConhol Toolbox
J
8. LMI Conhol Toolbox
,
< D 9. MappingToolbox
.
10.MATLAB Compiler
J
J Etc.andmanymore.
o
D
1..4.3 SIMITLINK
)
SIMULINK, developedby The Math Works, is a commercial tool for modeling,
)
-' simulatingand analyzingmultidomain dynamic systems.Its primary interface is a graphical
:' block diagrammingtool and a customizableset of block libraries. It offers tight integration
) with the rest of the MATLAB environmentand can either drive MATLAB or be scripted
-t from it. SIMULINK is widely used in control theory and digital signal processingfor
-t
multidomainsimulationandModel-BasedD esign.
?
q
e
q
,
t
)
D
t
=
Introduction Chapter1
-
-
-
I.4.4 WTIAT IS A SIMI.TLINK FUNCTION?
- In a State flow chart, a SMULINK function is a graphical object that you fill with
-
-
SIMULINK blocks and call in the actionsof statesand transitions.This function provides an
r<l
- efficient model design and improvesreadability by minimizing graphical and nongraphical
t: objects.Tlpical applicationsinclude:
It,
f'\
It 3 Defining a function that requiresSIMULINK blocks, suchaslookup tables.
lt o Schedulingexecutionof multiple confrollers.
It
l'-a
It
L:
lt
I
H
It
tl
1--
.J
l--
lJ.
IJ
IJ
E
k
E
F
E
l{
E
H -_6
t\
tl
tt
F- r
L-.
!:
l-
:\<
I
l\-
[:
lt-.
IL_- HA E R2
L-.
L--
LJ
LJ
LJ
tj
L:
L:
L:
PROPORTIONAL-
t:
l-
INTEGRAL.
r -'
tL:
l--
l{
DERIVATIYE
k
t
lj
CONTROL
u &
E
LEAD-LAG
COMPEI{SATOR
F
F
N
F
I
I
easily understood
I industries.It has been successfullyused for over 50 years. It is arobust
varied dynamic
! algorithm that can provide excellent control performance despite the
I of processplant.
characteristics
a
7
, Theselecturenotes:
;
,
. introducethe Proportional-Integral-Derivative (PID) control algorithm.
a
)
the algorithm'
t
) " discussthe role of the threemodesof
{
, . highlight different algorithm structures'
a
t
a
nDiscussmethodsthathaveevolvedoverthe last 50 yearsasaidsin controlloop tuning.
-
,
of approachinga
t After completionof this sectionof the coursea studentshouldbe capable
knowledge to
, loop tuning problem in a competentand effrcient manner and have sufficient
a
effectivelytune a PID control algorithm'
t
{
rt' 2.2 DIFFERENET TWES OF CONTROLLER
,
rt
u
t 2.2,I PROPORTIONAL CONTROL (GAIN)
{
J
The first elementof pID controlto be developedis Proportionalcontrol. The equation
e
-ra is simple:
O
r!
J
- setpoint (directaction)tll
error: measurement
:
\ or
J (reverseaction)tu
error: setpoint - measurement
t an increasein
:t Note the action may be either direct or reverse.In a direct acting control loop
element.
theprocessmeasurementcausesan increasein the output to the final control
It ll
t TheproportionalonlYequationis:
t i
{ OrSput= gain x effor + bias
I
't (such as Foxboro
The bias is sometimesknown as the manualreset. Some contol systems
't products,use proportional band rather than gain. The proportional band and the
gain are
:t relatedby:
T
b
\r
t-
t-
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
TED?O
Gain = Proportional Band
LEDTz
ProportionalBard = Effi-
2.1111
With proportionalonly control the confroller will not bring the processmeasurernent
to the set point without a manual adjusfrnentto the bias (or manual reset) term of the
equation.In the early days of contrpl the operator,upon observingan offset in the control
oresetting"the controller (adjustingthe bias).
loop would correctthe offset by manually
Rather than to require that the operator "manually reset" the control loop whenever
there was a load change control functions were developed to "automatically reset" the
conholler by adjustingthe bias term wheneverthere was an eror. This "automaticreset" is
alsoknown simply as "reset"or as "integral".
, .. .- t
t ^
6
I
-
t,
E-
l-
\.r Proportional-lntegral-Derivative
Control& Lead-LagCompensator Chapter2
tq
-
=
1
-3 I
I
: I
I
-: I
I
I
I
t I
I
I
t I
I
I
t I
I
t
t
M+;;;r;A
t
Variabl+
t
i<f
\{t
- Figure2.1:
- Theequationfor PI controlis:
.D
-
= O ut:gx KrxJ e d t
23trJ
:
.D
out: gainx (error+ integral(error)dt)
a
a
<l
The amount of reset used is measuredin terms of "reset time" in minutes or its
ts - inverse,"reset rate" in repeatsper minute. The following test can be performed on a
1O
-r{ dueto the gain"The output will continueto changeat a constantrate. The time is measured
D
\ from the time of the initial changeuntil the time that the instant changeis repeatedby the
f
J constantmovement.The repeattime, or resettime, is the time it takes for the reset effect to
f
t
t9
- t
o
Proportional-lntegral-Derivative
Control& Lead-LagCompensator Chapter2
repeat (or move the output the sameamount as) the gain effect. Its inverse is reset rate,
measuredin repeatsper minute.
Reset effect
-
t Gain effect
t
t
i
t
t % +1 0
t Eror 0
-1 0
Time
(-Tt-*
1 "Repeat- time
Figure2.2:
- The derivative term looks at the rate of changeof the input and adjusts the output
t
,' basedon the rate of change"The derivativefunction can either use the time derivative of the
, error,which would include changesin the set point, or of the measurementonly, excluding
t setpointchanges.
t
O u t : g x Kd xd e /d t
:
|l
t
2. With a constantoutput from the signal generatorand the controller in manual, adjust the
; setpoint to be equalto the input from the signalgenerator.
'f
! 3. Placethe controllerinto automaticmode.
:
: 4. Startthe ramp.
ra
t
% +1 0
-
a O
Eror
? -1 0
?
O 4--ur----*
? De r iva .tive tir n e
J
a Figure2.3:
o
J Note that when the ramp is started,with no derivative (dashedline) the output ftImps
\
f
up dueto the changein input and the gain. Using derivative (solid line) the outputjumps up,
i"'t risesin a ramp, and thenjurnps down. The differencein time betweenthe solid line and the
ir
rt dashedline representsthe amountof derivative,in units of time (usuallyminutes).
l-
F
l-t
l-t
]-t
I
lu
i;
Ir
L--
LD
L- /l
IJ 11
a^ t,
-l
Proportional-lntegral-Derivative
Control& Lead-LagCompensator Chapter2
Where
G: Gain
D = Derivative(minutes)
Showngraphically:
Manual
H+set
Ouput
Figare2"4:
Note that in the equationthe gainis multiplied by all threeterms.This is important for
thePID equationto be ableto be tunedby anyof the standardtuningmethods.
12
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
,' 2.3.1.APPLICATION
,'
t Lead-lag compensatorsinfluence disciplines as varied asrobotics,satellite control,
t automobilediagnostics,laserfrequencystabilization,and many more. They are an important
.t
buildingblock in analogcontrolsystems,andcanalsobe usedin digital control.
.t
Given a conkol plant, desired specificationscan be achieved using Compensators.I,
D, PI, PD, andPID, are optimizing controllerswhich are usedto improve systemparameters
suchas (steadystateerror, reducingresonantpe*, improve systan responseby reducingrise
time).All theseoperationscanbe doneby Compensators
aswell.
2.3.2THEORY
Both lead compensatorsand lag compensatorsinkoduce a pole-zero pair into the
openloop hansferfunction. The kansferfunction canbe written in the Laplacedomainas
v s -g
X .s-p 2.5121
v (u-"t)(t-tr)
X (s -pr)(s -pz)' 2.6421
tt
13
t,
L.- '!
f,
L: Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
t:
l-
\r
and stability of the system.The lag compensatorprovides phase
enhancesthe responsiveness
l"
\.
l" lag at low frequencieswhich reducesthe steadystateerror.
f{
t- The preciselocationsof the poles and zerosdependon both the desiredcharacteristics
t: of the closedloop responseand the characteristicsof the systembeing controlled. However,
t-
t\a
the pole and zero of the lag compensatorshouldbe closetogetherso asnot to causethe poles
L:
lL.-z
to shift right, which could causeinstability or slow convergence.Since their purpose is to
l.r
L
affectthe low frequencybehavior,they shouldbe nearthe origin.
lt
l,
-,
lr
L-
2.3.3 IMPLEMENTATION
L:
1-: Both analogand digital control systemsuse lead-lag compensators.The technology
f" used for the implonentation is different in each case,but the underlying principles are the
The transferfunction is rearrangedso that the output is expressedin terms of sums of
surme.
i>
E
L-.
l:
termsinvolving the input, andintegralsof the input andoutput. For example,
u
signalsampliff the noiseon the signal, since even very small amplitude noise has a high
derivative if its frequencyis high, while integrating a signal averagesout the noise. This
makesimplementationsin termsof integratorsthe most numerically stable.
E
E
H t!
f,
L4
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
electrical responseof this network to an input signal is expressedby the network's Laplace-
J
domaintransfer function, a complexmathematicalfunction which itself can be expressedas
t
J one of two ways: as the Current-gainratio hansfer function or as the Voltage-gain ratio
t transferfunction. Rememberthat a complexfunction canbe in generalwritten asF(x) : A(x)
J
+ iB(x), wherel(.r) is the "Real Part" and^B(i)is the "Imaginary Part" of the single-variable
t
functionF(x).
The "phaseande" of the networkis the argumentofF(x); in the left half plane this
is tan- t(f(") / A(x\).If the phaseangleis negativefor all signalfrequenciesin the network
then the networkis classifiedas a "lag network".If the phaseangleis positive for all signal
frequenciesin the network then thp network is classified as a "lead network". If the total
network phase angle has a combination of positive and negative phase as a function of
\
t frequencythen it is a "lead-lag network". Dependingupon the nominal operation design
t
parametersof a system under an active feedback contol, a lag or lead network can
t
causeinstability andpoor speedandresponsetimes.
t
t
15
,
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
2.4.L.I DEFINITION
t: of the system.utu
response
l-
{
J 2.4.I.2 CONTROL TTIEORY
J
t
\ In control theory, overshootrefers to an oubut exceeding its final, steady-state
t
\i
lr
Lra
(lnPo)z
Ilh'
lt
l-
?rs+ (lnPO)? 2.10t31
lr
_l
1-
l.
l{
l -l
L-
It
L-
tt
--
.J , 1 16
-, - -". -- (,
l^
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
-, For our exampleof 6': I secondandfor a 1 volt input stepwe havethe following:
1" 1
1
t I
t .6
t
at .7
t gt
6
ct .E
t
o
(J
6
.5
trL
(E
,L)
.4
.3
.2
.l
Figure2.5:
This frequencyis uzually referredto asthe 3 dB bandwidthor the half power bandwidth"
OR:
fl
L7
t,
!=-
L
L: Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
L:
L: The figure 2.5 often is quotedwhen characterizingthe performanceof oscilloscopes.
l..r
i'-a A 'scopewhose bandwidthis 100MHz and whose step responseis that of a simple RC
l-
network will have an internal rise time of 3.5 ns. Therefore,a pulse with rise time shorter
L:
L: than this cannotbe measuredon sucha'scope and it is fruitless to attemptto measurethe rise
l-
lra
l_/
time of a pulse if it is shorterthan that of the 'scope's.Even signals whose rise time
l\-
approachesthat of the 'scopewill be distorted so that their measuredrise time is inqreased.
LJ
LJ The measuredrisetime is givenby:
Ll
L: ftmeasured: {(ttr"on")' + (rtporr")t}% 2 " 1 { 31
lj
LJ Rule:if you want to measurethe rise time of a pulse;know what the rise time of your'scope
5
ir 2.4.2.I APPLICATION
E
L:
]j
In control theory, rise time is commonly defined as the time for a waveform to go
of its final value.(Nise2008)
from l0% to 90o/o
L: : 2"2sof- o.o7s(
t, "{do + 1.12 2.t2t3l
t:
L: where( is the dampingratio andcoois the natural frequencyof the network.
1
t, 'c,.16:
{L -C2 2.t3t3l
H
H
where( is the dampingratio andco6is the naturalfrequencyof the network.
E
E
H tt
f,
18
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
R i s e Ti me D1oop
rT1--.
Baseline
dershoct
J
representative pu lse '*,rraveform
Figure2.6
Ts = alKa): 4T
Where,T {ime constant 2J4t41
(= dampingratio
co: naturalfrequency
#19
t
b
I
I
\a
t PUT:
La DIGITAL CHANGE
,
ANALOG STEP
-
;
UTPUT
4
t PONSE
D
,
D
)
, SETTLING TIME..-..-.->I
t
Figore2.7:
2.4.3.LAPPLICATION
Settlingtime dependson the systemresponseandtime constant.
j ln (toleran ce fr action)
T*: -
= du*prttg ratio*natural frec1 2.15t51
J
lnf0.02) 3.9
r s" - - 1 ; ; - =T
(tJ", (U", 2.16rsl
J
)
tl
.-.,,l 2o
t,
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
2.4.4.L APPLICATION
2.8
Figure
For the examplesystem,the controlled system- often referredto as the plant - is a first order
systemwith a transferfunction:
G(s): Cdc(st + l)
. G d .:l
. t:1
. K :1.
. Kn canbe setto variousvaluesin the rangeof 0 to 10,
. The input is always l.
Here is a simulationyou can run to check how this works. In this simulation,the
system being controlled (the plant) and the sensorhave the parametersshwon above.
You can adjustthe gain up or down by SYausing the "-rowu buttons at bottom riglrt. We
can also enter our own gain in the text bon and seethe respons€for the gain you enter'
L
q proportional- Integral-Derivative Control & Lead- Lag Compensator Chapter 2
:
a:
-:
= Enterthe OC
= Galnforthe
Systemin rke
= baxat fte righl.
= Then clic* this
= txrtlon.
t -
ls
t
Closedhop
tt ffi gain = o.s
lrr
, g
Cls$sdhsp
IO TimeSon*t =o.s s€c.
t,
,$
ts
'"t
ir
Figure2.9:
ir
H
L: 2.5 IMPORTA}'ICE OF PID A}ID LEAD-LAG COMPENSATOR
f=
r\e
ld OVER TIME RESPONSE SPECIFICATIONS
l'-1
Ib-r
L: 2,5,L INTRODUCTION
ld
This tutorial will show the characteristicsof the eachof proportional (P), the integral (I)'
I_
IJ and the derivative (D) controls, and how to use them to obtain a desired response-In this
It
L
tutorial. we will considerthe following unlty feedbacksystem:
l!
lrt
L
t:
i-t
E
I
Ea
L;
l:
l^ Plant: A systemto be controlled
F
--
1:
l--
lr t
L
r---l
lr
Control& Lead-LagCompensator
Proportional-lntegral-Derivative Chapter2
Controller: Provides the excitation for the plant; Designed to control the overall syatem
behavior.
ar K, , EosI+Kos+K,
.8.- +r
Ptr's ",
+l},^i
2Jg.2l
. Kp: Proportionalgain
. KI: Integralgain
. Kd: Derivativegain
First, lefs take a look at how the PID controller works in a closed-loopsystemusing
the schematicshown above. The variable (e) representsthe hacking eror, the difference
betweenthe desiredinput value (R) and the actual output (Y). This error signal (e) will be
sentto the PID confioller, and the conholler computesboth the derivative and the integral of
this error signal. The signal (u) just pastthe contoller is now equal to the proportional gain
(Kp) times the magnitudeof the errorplus the,integralgain (Ki) times the integral of the emor
plus the derivativegain (Kd) times the derivativeofthe eror.
u : K'B+Krfedt K"#
- 2.lgt2l
This signal (u) will be sent to the plant, and the new output (Y) will be obtained.This new
output(Y) will be sentback to the sensoragainto find the new error signal (e). The controller
takesthis new error signal and computesits derivative and its integral again; This process
goeson andon.
tl
23
t,
Proportional-lntegral-Derivative
Control& Lead-Lag Compensator Chapter2
. Table2.1
,r
_, _ -- 124 +, ,
II
tt
.-.
CHAPTER J
MATLAB
PROGRAMING
&
SIMULINKMODEL
OF ARVIATURE
CONTROLLE,DDC
MOTOR
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
3.1 INTRODUCTION
The servo motor is actually an assenrblyof four things: a normal DC motor, a gear
reductionunit, a position-sensingdevice(usually a potentiomsfsl-4 volume conhol knob),
and a control circuit. Dc servo motors are normally used as prime movers in computers,
numerically conholled machinery, or other applications where starts and stops are made
quickly and accurately.
Where,
J:Total momentof Inertia of the rotor togetherwith that of the reflected load on the
rotor side.
B=Total viscousconstant.
KFMotor torqueconstant.
Ra=Armatureresistance.
La:Armature inductance"
,
I
J
tt
26
1'
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
J: 5.0*10 - 4kgtn2/sec2.
B 6.0*10^-3 N-m/sec.
at
t Kt 0.085N-m.
t
t Kb 0.lV/rad/sec.
Ra 5Cr
La negligible.
t
27
a
z
f.-
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
3.3.1 RESI,]LT
lj
H
lj
IJ
u
E
E Figure3.1
The responseof the motor position will found to grow continuouslywhen the unit
ste,pvoltageis appliedto the armatureterminals.The open loop systemis acting as a pure
integratorso there is a needfor feedbackthat will stabilize the systemfor a stepinput. The
closedloop fransferfunction of the systemwith unit feedbackcanbe obtained-
h:1
gtb=feedback(g,h);
step(gtb,t)
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
t\
29
t,
2
a
:
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
:
.-
e 3.4.I.RESULT
.a
a
!
:
.-
,
D
t
D
,
t
t
{
rq
a,
i{
=
!
-t
-:
,
t
Figure3.2
,
r.-
t
In this aboveprograming,we usea feedbackwhosegain is I and it is known asunit
, feedback.Theopenloop systemactslike anintegrator.So this feedbackmakesthe openloop
t systeminto a closeloop systernwhich will stabilizethe system.The step responseof the
t
aboveclosedloop systemis shownin thenextpage.
t
:r
Though we use a closed loop systern,the responsewhich is obtained from the
-t
! previousplot is not desirable.Here, we will designa position control systernwith unity
=t
feedback.The designspecificationsfor a stableanddesiredsysternareasfollows.
-a
I 1. Steadystateerrorin angularposition,0ess:0.
F
I
l{ 2. Settlingtime,ts:0.2 sec.
t-
L{
I J
3. Overshootmustbe equalor lessthan5%.
I
li.a
l- From the closed loop response we find that the settling time is
L.-
is much morethan the desiredvalue of 0.2 sec,eventhoughthe
1.55(approximately),which
L:
lr
overshootis O.Sincethe rising time is muchhigh, so to reduceit we usedthe PID controtler"
It is to be mentionedthat in majority casesa PI or PD controllermay be adequate& there
lr
t,D
It t30
T
I'
MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
t
p may not be needto usethe PID controller.However,full PID controllermay be designedfor
p better performance.In order to improve the settling time, we chooseanotherresponseas
p follows.
p
p
t 3.5 PROGRAMING FOR STEP RESPONSE OF A CLOSED
)
LOOP SYSTEM USING PID CONTROLI;ER
)
)
) iI=5. 0e-4 ;B=5. 0e-3 ;Kt=0. 69$ ;Kb=0 - L ;
) Ra=5;
b num=[ (Kt) / (Ra*iI) ]
) den= [1 ( (B/.]) + (Kb*Kt) / (Ra*,J) ) 0l
?
p KP=3oo
tKd=O.5
p tKi=O.01
t nr:m1= tKpl
p *num1 = [Kd Kp]
t *num1 [Kd Kp Ki]
t denl= t1l
?
*den= t1 0l
D
num2= conv(numl-,num)
D
den2= conv(den1 ,den)
?
? g1=tf (num2 , den2 )
? t= 0:0 .005:1-
? 92=feedback (91, 1 )
? step (92, t)
?
?
't
a
-t
!t
b
it
i-t
i't
1..
l-
!-r
It
tsr
lr
lr
1.\
b tt
t:
tr
I
t,
31
>\
ar-
a
1! MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
:
{ J 3.5.1"RESI,]LT
{U
t
a
t
t
?
?
?
?
?
t
a
:t
?
.t
5
{
t
-t
?
ta Figure3.3
-
-
E-
J
From the previousclosedloop responsewe find that the settling time is 0.502 sec,
:1
which is not desirablevalue.It is morethanthe desiredvalue of 0.2 sec"Full PID controller
:3
may be designedby trial and errorprocedure.Now changingthe valuesof Ki, Kd & Kp we
iD
i-r getanotherclosedloop responseasfollows.
i"
f-t
it
it
L-O
L
L
LI
L:
L:
L:
L:
L
t-
l- tl
It 32
TJ
It
2
a
: MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
:
: 3.6 PROGRAMING FOR STEP RESPONSE OF CLOSED
: LOOP SYSTEM:
!
a
t ;Kb:0"1;
J:5.0e-4;B:6.0e-3'KF0.085
{ Ra:5;
-
-J num:[(Kt)/(Ra*J)]
:
den:[l ((B/J)+(Kb*Kt)/(Ra*J))0]
D
Knl
t
? Kd:20
? %Ki=0.01
J %numl: [Kp]
t numl : [Kd Kp]
?
? %onuml[Kd Kp Ki]
t 7odenl:[]
t denl: [l 0]
{
num2: conv(num1,num)
{
den}: conv(denl,den)
-t
? gl:t(num2derZ)
? F 0:0.01:2
-t #=feedback(g1,1)
? step(g2,t)
?
?
?
?
:'
!
a
-t
5
-{
a
€
=
a:
t
t
t
t) 33
I t
l.
t
t
t MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
t
t 3.6.1 RESLILT
t
t
t
t
t
t
I
I
I
t
t
)
t
t
,
t
t
)
t
t
I
I
Figure3.4
t
t From the previousclosedloop response,we find that the settlingtime is 0.433 sec,
t which is undesirablevalue. In a control systemresponse,the time constantis directly related
t to the settlingtime. A decreasein the value of the time constantof a control systemindicates
t thatthe steadystatereachesearlier. Therefore,lower the time constanti.e. lower the settling
t
t time fasteris the time responseof a control system.So, we choosethe value of Kp, Kd and
t Ki againand createanotherresponseusing PID controller for the optimal output responseand
a the settling.The new responseis shownbelow.
t
t
a
a
a
t
{
{
-
J
-
D
7 , - ----"--{
t
34
a
zt
l-; MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
{
.'
3.7 PROGRAMING FOR STEP RESPONSE OF A CLOSED
= LOOP SYSTEUI USING PID CONTROLLER:
{
!
! J=5. 0e-4; B=5 . 0e-3;Kt=0 " 085;Kb=O.1;
'3 Ra=5;
r.a
T
num= [ (Ktr) / (Ra*,J) ]
T
den= [1 ( (B/\T)+ (I{b*Kt) / (Ra*J) ) 0]
f
Kp=350
it
\q Kd=50
rt
Ki=300
It
\* tnuml= [Kp]
lf tnuml = [Kd Kp]
{
I' numl = [Kd Kp Ki]
it tdenl=
i-t
I
tll
S denl= tl- 0I
num2= conv(num1 ,num)
it
sl- den2= conv(den1 ,den)
g1=tf (num2, den2 )
it
i-3 t= 0 :0 .01-:2
It 92=feedback(91, 1 )
f step (92, t)
<
L-
t
|J
ie
It
i3
i3
l-
-
it
it
5
-
it
it
L;
-
l-
h{
lD
lr
lr
l.D
t
lr t
I
?E
er
Lo
I
-t
2
.3
r- MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
r-
L.
3.7.1 RESLILT
r- -
=
:3
'!
-3
q
-
f
t
I
f
D
D
€
-3
iq
It
€
!
!
-3
It
f
Figure3.5
{ tf,
t
J
-qa, In this program,we increasethevalueof Kp, Kd andKi.We increasethe proportional
U control Kp to improve the rise time, add derivative action Kd and increaseit to decreasethe
:3 overshootand use an integral Ki to decreasethe steadystate error. The stepresponseof the
:r aboveprogramis shownin the next part.
{
{
In the previousresponse,it is obtainedthat the settlingtime is0.00985sec.which lies
'3
'3 in the desirablerangeof it. The overshootis 0.075%which is also acceptable.
The iterative
-t designprocedureof a PID controlleris illustratedin MATLAB environment"As in many
{ casesof systemdesignthereis no uniquesolutionto the designproblemof a PID controller
a .Howeverthe final choiceof the PID parameters
may be dictatedby hardwareconskainsand
-t
.a finally we obtaina satisfactoryresponse
of the system.
rq
It
!
,D
a
o
7
€
! MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
!
{ry
3.8 SIMT]LINK MODEL
! ln the SIMULINK model, we use three tlpes of controller. These are-l) Proportional
t controller 2) Derivative controller and 3) Integral controller. trnthe following model, these
t three controllers are connectedin cascadeform. The gain I is used for the proportional
t conholler, where as the gauin2 is the derivative controller and the gain 3 is for the integral
t contoller. Actually by using this three gain, we can select the value of the Kp, Kd and Ki
: respectively.
3
t
e
a
I
f
t
t
a
t
t
t
t
t
a
e
t
e
it
e
t Figure3.1:
t
t
t
t
t
t
t
t
t
t
t
t
-J
I
f
e t
3 V l-- - _.
o t
€
: MATLABPrograming& SIMULINKmodelof armaturecontrolledDCmotor Chapter3
r-
-
'.-t 3.8.1 RESI,]LT
r-
D
3
:
:
.-
D
t
D (l)
tA
I E
o
a
D .h
(u
D fr
a
-a
.a-,
t-
a
! Timein sees
!
q
Figure3.6
-
J
t3 Here we use threecontroller(Kp,Kd & Ki) in SIMULINK Model. By the use of this three
-t f controllerwe can achievethe aboveresponse.Here we can seethe responseis closelymove
-art
towardsthe stability"
?
?
?
?
?
{
a
!
!
{
a
a
€
a
-
-
t
rt
f
ft
? | 38 +_-_ _-
1.,
I
--t)
=
r-
r{ -
f
r-
:
r-
r>
€
-3 CH APT E R
'3
{
r-
t
-f
t
f
t
t
't
--
t-
-
-!
-t
-!
"3
<
U
It
f
OF
ID
rt
r
-.qt
U
J
THE PROIECT
-3
!
{
E
!-
!'l ^
_-
l'
4
t-
r-
-
l-
lu
lu
L:
It
lr
?
3 Futureaspectsofthe project Chapter4
3
.q
tr
This projectcould find its applicationin different fields,asfollows:
!
a
t 4.I DC POSITION CONTROL SYSTEM
{
q
It
rt
f
\
f
t
t
D
3
f
t
3
3
t
t
t . Figure4.1
.-
--a
la
t 4.L"L INTRODUCTION
q
-rt One of the most coillmon examplescovered in text books and literature on linear
3 systemsis a DC position control system).This systemis easilyunderstoodand has a second
t,
ra ordertransferfunction in the standardform, for which a well developedtheoreticaltreatrnent
t
is available.
!
3
3 This unit providesthe studentsan opportunity to study andoperatea practical electro-
t mechanicalangular-position-controlsystem. The system is built around a good quality
t permanentmagnet armature-conholledDC motor, speedreduction gear-set,potentiometric
t
error detector using special 360o revolution servo potentiometers,a tachogeneratorfor
t
q velocity feedback and associatedelectronic circuits. Unlike simulated systems, e.g. our
q, LINEAR SYSTEMSIMULATOR,thepositioncontrol systemnaturallyconsistsof non-ideal
e parametersviz. saturation of amplifier and motor current, dead zone and backlash, non-
t linearity in the motor and gears,imperfectionsin mechanicalfabrication and somewhat
{
q- uncertainorder of the complete system due to filters, various time constants and load
r:f
E
F3
systemsencountered
in practice.
F3
A difficulty which is facedwhile working with many practical control systemsis that
!
.! their responsesare rather slow (Note that in a simulated systeinthe common practice is to
ia valuesof forward gain. Study of the effect of velocity feedbackon the transientand steady
ie stateperformanceof the systernaswell asits liability.
it
i{
The experiments would involve calibration and operation of the waveform
it capture/displaysect)onas a first step,It may be mentionedherethat due to the non-linearities
ie
i'r andotherimperfectionsand uncertaintiesexisting in a physical systemof the presentt1pe, a
it quantitativeverification of the resultswith theoretical analysisis not recommended.This is
!.
I best done on a simulated system. Of course the experimental work does include
i{ determinationof rise time, overshoot,steadystate errors etc. for various conditions for an
i-r
S evaluationof the systemperformance.
I
lv
-
i:l
iil
iD
ie ,\
t
A' -Lr_
t
lo
!
: Futureaspectsofthe project Chapter4
:
e FEATURES & SPECIFICATIONS
:
!
t
!
.e I
U
\
U
D
i]
il
ir
f
! operationLiteratureandpatchcordsincludedEssentialaccessories
-a CRO.
!
€
!
a
!
{
?
:3
:3
.aU
?
?
?
?
:3
:t
! Figare4.2
{
't
't
q
{
't-
{
<U
-1It
-\
U
\r
rl|
-? fl
42
I 1J
7
b
-
lr Future aspectsofthe proiect Chapter4
-P
D 4.2 ROBOTICS
-
D
-a
t>
3 4.2.I INTBODUCTION
a A robotic arm is a robotic manipulator,usually programmablg with similar functions
!D
-
q
to a human aru. The links of such a manipulator are connectedby joints allowing either
f
rotational motion (such as in an articulated robot) or translational (linear) displacement.The
f
tinks ofthe manipulatorcan be consideredto form a kinematic chain. The businessend of the
f
\ kinematic chain of the manipulatoris called the end effector and it is analogousto the human
f
r\
r hand. The end effector can be designed to perform any desire.dtask such as welding,
D grrpping, spinning etc., dependingon the application. For examplerobot arms in automotive
f assemblylines perform a variety of tasks such as welding and parts rotation and placement
t during assembly.
3
In spacethe SpaceShuttle RemoteManipulator Systemalso known as Canadannor SSRMS
!-
rt
! and its succ€ssorCanafurm2are examplesof multi degreeof freedomrobotic arms that have
t been used to perform a variety of tasks such as inspectionsof the Space Shuttle using a
t specially deployed boom with camerasand sensorsattachedat the end effector and satellite
d
ra
- deploymentand retrieval manoeuvresfrom the cargo bay of the SpaceShuttle.
f
-t The robot arms can be autonomousor controlled manually and can be used to perform a
rD
-
rt variety of taskswith greataccuracy.
!D
q
U The robotic arm can be fixed or mobile (i.e. wheeled) and can h designedfor industrial or
J homeapplications.
ra,
I
ral
t
rq
It
ra
U
!
q
!
{
{
{ Figure4.3
q
!
{
r{r
!y
ra
J
trf
rt
f
'r3
I
7
tr
L. Futureaspectsofthe project Chapter4
i-
\-
-
La
4.2.2 DGERIMENT
ra
-
>
In order to implement a force conhol schemeinto the systemseveralexperiment is
!
! carriedout to achievethe key requirementof the project. The first experimentinvolves the
! operationof the robot systemin free spacemotion. Severaltestswere performed in world
ra
ID
r-, coordinatesin x-axis direction where the force is applied to the force/torquesensorand the
:
data measuredis then routed to the robot controller. This generatesincrrementalposition
!
f
demandfor the robot system and the robot arm will move at a velocity proportional to the
t applied force. Determination of specific gain value for a conventionalforce controller in
Lr'
f principle is relatively straightforward.However, the problem ariseswhen the specific value
f
ofKis unknownor variable.Furtherexperimentmust be doneto determinethe ideal Kvalue
f
to avoid any serious consequencesin practical robotic systems particularly in system
t.q
It stability.A secondexperimentwill be carriedout using conventionalPID control for force
rqt
J conholler to improve stability performanceand manoeuwability of the manipulator arm in
! free spacemotion test as well as experimentin contactwith the environmentwhich is a key
!
requirementof this project.This will be closedloop conhol systern.The force conkol loop
t
!
compensates
the difference or error betweenthe desired and actual force and then send the
-
r-
: desiredposition to the robot conholler. Further experiment will be done to develop force
f controlschemefor TMS robotic systerrbasedon hybrid position/forcecontrol scherne.
\
f
r-
!D
rt
!
hr
f
D
rt
?
q
t
?
.lt
q
q
!
{
!
.a
ID
rq
U
D
f
7 #44
tl
t t,
t Future aspectsofthe project Chapter4
E
t 4.9 RADAR CONTROL
T
T 4.8.I INTRODUCTION
T Radar is an objectdeteotion systemwhich useselectromagneticwave$-specifically
I radio waver-to determinethe range, altitude, directioq or speedof both moving and fixed
t
t objects such as aircraft, ships, spacecraft, guided missiles, mofor vehicles, weather
t formationg and terrain. The radar dish, or antenn4 transmits pulses of radio waves or
? microwaves which bounce off any object in their path. The object returns a tiny part of frre
T wave'senergyto a dish G ant€nnawhich is usually locatedat the samesi& as tlre transmitter.
;t
?
\
)
\
;
t
l
rI
It
p Figure4.4
) 4.8.2 THE SYSTEM APPROCH
)
t\ The main problem of Radar system is in different areas'liks prgvision a$curat€
) measurementsof position using radar, improve methods of manual tracking remote power
t
Ir control of gun positioning improve predirector design, automatic fuse setting etc. the
:
t eventualaim is to have fully automatichacking of tho target and aiming of the guns.
t
I The biggostsontrol systemrequirernentat the beginningof the war was the provision
t of a fast, accurate,high powered position control system that can be used to move guns,
t searchlightsand radar antennae.To meet these requirementtechniquesfor the analysis and
t design of contuol systemhas to be refined and developed,and a whole range of imprcved
tp components.-motors, amplifiers, transducer has to be developed. There is also other
problems:the designof fitters to smooth signalsobtain from manualtracking and radars,the
' design of circuits to give dre best estimateof the ftturc position of a targeq techniquesfor
analysis of servomechanismswith pulsed data and for the design of system with optimum
responsewith respect to some dynamic performancecriteria in the presenceof noise and
I di*urbance.
I
)
)
p
t
)
CONCLUSION
Conclusion Chapter5
6.1 CONCLUSION
The processeswhich are taken in this report for MATLAB programing are not
completelyoptimum. Someapproxirnationare taken to completethe report and the value of
Kp,Kd and Ki are chosenby hial and error method.Throug[r,the results which are obtained
from the responsesarein the desirablerange.
tt
# 47
t,
6. REFERENCES
1. Equations:
t!
t-
t€
-r
l=
F'
L.
L.
ta
ti
L:
t:
t:
lv