Digital Control Systems
Uploaded by
cooperroverDigital Control Systems
Uploaded by
cooperrover = 200 are selected. The plot of Lm 1/H:(jw)He( jw) $
—54 dB, where |C(jw)/D( jw)| ~ |1/He( jw)He(jw)| over the range 0 < w < we =
10 rad/s, is shown in Fig. 2.30. This requires that 200K. > 50 x 10’. The open-loop
transfer function of Fig. 2.26 is now
200K -(s + 4)(s + 8)
s*(s + 1)(s + 200)*
(Ke /1250)(0.25s + 1)(0.125s + 1)
* 35 + 10.0055 + 1) 281)
G2(s)H2(s)He(s)
"
The poles and zeros of Eq. (2.81) are plotted in Fig. 2.31. The value of static-loop
sensitivity 200K. = 5 x 10° yields the set of roots
pi = —3.8469 pz = —10.079
p34 = —23.219 + j66.914 ps6 = —270.32 + j95.873
and are shown in Fig. 2.31. This choice of gain results in a satisfactory design; ie.,
Mp = 0.001076 < 0.002 (—54 dB), tp = 0.155 s (<0.2 8), c(t)ss =0,and0 Sw Supe
10 rad/s. Note that the gain can be reduced until either M, = 0.002 with tp < 0.2 s or
tp = 0.2 s with Mp < 0.002 is achieved. Although it is not specified in this problem, the
phase-margin frequency we = 56.2 rad/s must also be taken into account in the design
of a practical system. 2.9 DISTURBANCE REJECTION 6]
Ss
a
ae
&
a
Lm(i/H,H,]
40
20
ok
20k
49h
—60}—
—so
—100
dB
1000
0.1
()
Bode plot for Example 2.5: (a) Straight-line approximations; (b) exact plots.
FIGURE 2.30 62. CONTINUOUS-TIME CONTROL-SYSTEM RESPONSE CHARACTERISTICS
Lt ! a !
-300 -280 -260 -240 -220 ~200], -180 -160-100 -80 -60 40
FIGURE 2.31
Sketch of root locus for Eq. (2.81) (not to scale).
2.10 SUMMARY
This chapter has presented the standard FOM based upon a simple second-order
continuous-system model. The qualitative correlation among the frequency response,
the root locus, and the time domain for second- and higher-order systems is presented
as an aid in achieving the desired system performance characteristics. That is, the
FOM in conjunction with this correlation are used as a guidepost for evaluating a
system’s performance. A method for adjusting the forward loop gain of a basic unity-
feedback control system to achieve a desired value of Mj, by use of Nichol’s chart is
developed. When the desired FOM cannot be achieved by the basic system—by mere
gain adjustment—then one of the compensation design procedures, discussed in this
chapter, may be employed to try to achieve these FOM. A discussion of synthesizing
desired tracking- and disturbance-control ratio models is also included. These models
are utilized in many compensation design techniques as an aid to achieving the de-
sired FOM. In the following chapters the controller design approaches presented in
this chapter are applied to sampled-data systems. CHAPTER.
3
LINEAR
SYSTEMS AND
THE SAMPLING
PROCESS
3.1. INTRODUCTION
The mathematical techniques used in this text are based upon the model of linear time-
invariant (LTI) systems. Therefore, unless stated otherwise, from here on only LTI
systems are considered. Section 1.2 introduces the concept of the sampling process,
as illustrated in Figs. 1.2 and 1.3. As presented, a sampled continuous signal yields
a signal in the form of discrete values. This chapter develops the sampling process
in more detail and introduces the reality of sampling-time selection. Also, it presents
time- and frequency-domain representations of sampled-data systems in terms of linear
equations, as well as techniques for continuous-signal reconstruction from a sampled
signal by means of a data-hold device. In sampled-data control systems, as shown in
Fig. 1.7, where at least one component operates in continuous time, it is necessary
to include a data-hold device to maintain control over the system between sampling
instants.
3.2, LINEAR TIME-INVARIANT (LTI)
SYSTEM
A system is said to be linear if and only if the superposition theorem can be satisfied.
This theorem states that “the output response of an LTI system to a number of si-
multaneously applied inputs is equal to the summation of the system responses when 64 LINEAR SYSTEMS AND THE SAMPLING PROCESS
ra cy (kT) (kT)
|) BtkT)
FIGURE 3.1
Single-input-single-output_ (SISO)
system: (a) Continuous time; (6)
discrete time.
(a) )
each input is applied individually.” For example, consider a single-input-single-output
(SISO) system, as shown in Fig. 3.1(a), whose output is c(t) = ¢;(t) due to an input
r(t) = r)(t). This system is said to be linear if for the input r(t) = a)ri(t) + a2r2(t)
the output is given by
e(t) = aycy(t) + a2c2(t) 3.1)
When a; and a2 are constant real numbers, and c; is the respective output due to ri,
in terms of function notation, the composite output may be expressed as
= flrl= flan + aorel = ai fLfil + af fra] (3.2)
The notation f represents the manner in which the system function g operates on the
input signal or transforms the input signal to yield an output signal.
For the system whose input and output quantitics represent discrete or sampled
quantities, then Eq. (3.2), using the superposition theorem, can be expressed as
(kT) = flarri (kT) + aara(kT)] = a1 flrs(kT)] + anf fr2kT)] 3.3)
The results for a SISO system may be extended to a multiple-input-multiple-
output (MIMO) system having m inputs and p outputs. The m x 1 input column vector
r and the p x 1 output column vector ¢ are expressed, respectively, as
TM C1
T2 C2
r=|. c=
Tm Cp
where r; and cj are the elements of the respective vectors. Therefore, for either a
continuous- or a discrete-time system to be linear, the relationship
= flair) + a2r2) = ai f(t) + offre] (3.4)
where a; and a2 are constant real numbers and r; and r2 are any two input vectors,
must be satisfied. If this relationship is not satisfied, the system is said to be nonlinear.
Example 3.1. The following first-order differential equation describes a system having
the input r(¢):
c(t) = r(t) +a (3.5)
To determine if the system is linear, proceed as follows:
Solution. Consider r(t) = airi(t) + a2r2(t). 32 LINEAR TIME-INVARIANT (LTH SYSTEM 65
Step 1. For ri(t).
canta (3.6)
Step 2. For ro(t), ,
aQ=nta (3.7)
Step 3. For r = air) +a2r2,
c=artar2+a (3.8)
Step 4. Substituting Eqs. (3.6) and (3.7) into Eq. (3.1) yields
c= air) +azrr + (ai taza (3.9)
Step 5. As Eqs. (3.8) and (3.9) are not identical, assuming a + a2 4 1, Eq. (3.5)
is not linear.
The system is nonlinear since Eq. (3.4) is not satisfied.
Example 3.2. An approximation for numerical differentiation [see Eq. (3.33)] of c(t) =
F(t) is given by the discrete-time equation (first-order backward difference)
1
(kT) = = {r(kT) — rk — T}} (3.10)
where T is the constant sampling time of r(t) and k = 0, 1, 2,.... Is Eq. (3.10) a linear
equation representing a linear system?
Solution
Step 1. For ri(kT),
c(kT) = FlMAT) ~ rik ~ DT} (3.11)
Step 2. For r2(kT),
ek) = Ff (rx(K) ~ lth ~ DT) G.12)
Step 3. For r(kT) = ayr\(kT) + a2r2(kT),
c(kT) = Firer) = rik - DTV} + Finan) ~rf(k — TY} (3.13)
Step 4. Substituting Eqs. (3.11) and (3.12) into the discrete version of Eq. (3.1)
yields
(KT) = ayer(kT) + azc2(kT)
FlrAKD) ~ rk = WTI} + Blk) — ral NT G4)
Step 5. As Eqs. (3.13) and (3.14) are identical, Eq. (3.10) is linear.
Example 3.3. To determine if (kT) = r°(kT) is a linear equation proceed as follows. 66 LINEAR SYSTEMS AND THE SAMPLING PROCESS
Path |
System
ra) e() = f[r)] t ey(f- t)
un T*] L POHS ER] delay PTT ete - 97]
{ System
! t [rt — = et - 7)
Le. —E —
| delay ik PFT FP gute — pT = elk - PT]
Path 2
FIGURE 3.2
Time-invariance concept.
Solution
Step 1. For ri(kT),
e(kT) = ri(kT) (3.15)
Step 2. For r2(kT),
ex(kT) = ri(kT) (3.16)
Step 3. For r(kT’) = airi(kT) + a2r2(kT),
e(kT) = abr}(kT) + 2ayayri(kT)r2(kT) + 0373(KT) @.17)
Step 4. Using the discrete version of Eq. (3.1) and substituting Eqs. (3.15) and
(3.16) into it yields
(KT) = ayer(kT) + anex(kT) = ari (KT) + a2r3(kT) (3.18)
Step 5. As Eqs. (3.17) and (3.18) are not identical, the given equation (kT) =
r*(kT) is not linear.
A system is said to be time-invariant when the coefficients of the differential or
difference equation relating the system’s output to its input do not depend upon time.
In general, if a specified input is applied to a given system at any time ¢; or t;, and if
the output responses to each application of this input are the same, the system is said
to be time-invariant. The concept of time invariance can be illustrated by using the
notation of a function with a time delay of r seconds and a block diagram as shown
in Fig. 3.2. The delay in path 1 acts on both the implicit and explicit time variables
in the equation for c(-). The delay in path 2 acts only on the implicit variable in the
equation for c(-). Thus, if the output quantities of path 1 are equal to the corresponding
outputs of path 2, the system is said to be time-invariant; i.e.,
et — 7) = fir -7)] for a continuous-time system (3.19)
e{(k — p)T] = f{rl(k - p)T]} for a discrete-time system (3.20)
where rT = pT and p=1, 2, 3,.... 3.2 LINEAR TIME-INVARIANT (LT) SYSTEM 67.
Example 3.4. To determine if the system described by
(t) + c(t) = tr(t) (3.21) ;
is time-invariant, proceed as follows.
Solution
Path 1.
a(t) + er) = tr(t) > a(t — 7) Felt — 7) = (t- Tr(t—7)
= tr(t—7r)-—7r(t—7) (3.22)
Path 2.
x(t —7)+e(t—7) = trit—7)
= ir(t—1)—7r(t- 7) +7r(t—7) (3.23)
As Eqs. (3.22) and (3.23) are not identical, c(t — r) c(t — 7). Therefore, the system
is not time-invariant.
Example 3.5. The solution for determining if the discrete-time system described by
c(kT) = kT r(kT) is time-invariant is as follows.
Solution
Path 1.
ex(kT) = kT (kT) > ei[(k — p)T] = ((k — p)T rk — p)T) (3.24)
Path 2.
ca[(k — p)T] = kT r{(k — pT}
= [(k — p)TWr[(k — pT) + pT rik — p)T) (3.25)
As Eqs. (3.24) and (3.25) are not identical,
elk — p)T] ¥ eal(k — p)T]
Therefore, the system is not time-invariant.
Example 3.6. Repeat Example 3.5 for the system described by c[(k + IT] — c(kT) =
(kT). As both paths 1 and 2 result in
of(k + 1 — p)T] - ef(k - pT] = r{(k — pT)
the system is said to be time-invariant.
From these examples it can be stated that a continuous (discrete) system is time-
invariant if the variable t(kT) does not explicitly appear in the system’s differential
(difference) equation, i.e., as a coefficient. All system models in this text are assumed
LTI owing to the general techniques used in control-system analysis and synthesis.
Also, most systems can be approximated over some range of operation as LTI. 68 LINEAR SYSTEMS AND THE SAMPLING PROCESS
3.3 SOLUTION OF LINEAR DIFFERENCE
EQUATIONS
Sampled-data control systems discrete variables can be modeled by linear difference
equations. In order to simulate or model the continuous-time variables on a digital
computer, it is necessary to develop the methods of discretizing continuous variables,
This section focuses on the use of a difference operator as an approximation to a
differential or integral operator.
Given the set of discrete values c(NT), c[(N — 1)T],...,c(0), where N =
1, 2, , one defines the first-backward difference V as
Ve(NT) = (NT) — c{(N — )T] (3.26)
where c(NT) represents the present value and c[(N — 1)] represents the value one
period behind c(NT) or “backward.” The second-backward difference V* is defined
as
WANT)
VIVe(NT)] = Ve(NT) — Ve[(N — IT]
e(NT) — ef(N — NT] — {el(N — 1)T)] - e{(N - 2)T]}
(NT) = 2cl(N = 1)T] + cf(N - 2)T] 3.27)
and the rth-backward difference V" is defined as
V'ANT) = V9O1e(NT) — VON — 1)T] (3.28)
where r = 1, 2, 3,.... The representation of a sampled-data control system by
difference equations permits its solution on a digital computer.
Example 3.7. Evaluate
t
a= [ er) dr (3.29)
0
by means of a numerical integration technique.! The integral of Eq. (3.29) is approxi-
mated (see Fig. 3.3) by
NT N-1 N-1
(NT) = [ e(r) dr) e(kT) At = So Teter) (3.30)
° k= k=O
where k = 0, 1, 2,...,At =, and N is a constant integer greater than zero. e(kT)
represents a constant function for the time interval kT
You might also like
- McGraw-Hill - Digital Control Systems - Houpis & Lamont75% (4)McGraw-Hill - Digital Control Systems - Houpis & Lamont768 pages
- Digital Control System - PhilipsNagle PDFNo ratings yetDigital Control System - PhilipsNagle PDF701 pages
- Houpis, Constantine H, Digital Control Systems Theory, Hardware, SoftwareNo ratings yetHoupis, Constantine H, Digital Control Systems Theory, Hardware, Software768 pages
- Sliding Mode in Control and OptimizationNo ratings yetSliding Mode in Control and Optimization173 pages
- Chen ActiveNetworkAnalysis - FeedbackAmplifierTheory PDFNo ratings yetChen ActiveNetworkAnalysis - FeedbackAmplifierTheory PDF880 pages
- Control Engineering - Theory, Worked Examples and Problems88% (8)Control Engineering - Theory, Worked Examples and Problems117 pages
- Lec#01 Introduction To Digital Control SystemsNo ratings yetLec#01 Introduction To Digital Control Systems10 pages
- Dr. Dong-Jin Lim - Control Systems Engineering - Design and Implementation Using Arm Cortex-M Microcontrollers-Independently Published (2021)100% (2)Dr. Dong-Jin Lim - Control Systems Engineering - Design and Implementation Using Arm Cortex-M Microcontrollers-Independently Published (2021)449 pages
- 978-0134407623 Modern Control Systems (13th Edition)pdf download100% (1)978-0134407623 Modern Control Systems (13th Edition)pdf download59 pages
- Full Download Modern Control Systems 13th Edition (Ebook PDF67% (3)Full Download Modern Control Systems 13th Edition (Ebook PDF41 pages
- Download Complete Modern Control Systems 13th Edition (eBook PDF) PDF for All Chapters100% (1)Download Complete Modern Control Systems 13th Edition (eBook PDF) PDF for All Chapters50 pages
- (eBook PDF) Modern Control Systems 13th Edition by Richard C. Dorf pdf download100% (2)(eBook PDF) Modern Control Systems 13th Edition by Richard C. Dorf pdf download55 pages
- Control of Digital Systems: Automatic Control and System TheoryNo ratings yetControl of Digital Systems: Automatic Control and System Theory54 pages
- El 3015 Sistem Kendali: Control SystemsNo ratings yetEl 3015 Sistem Kendali: Control Systems24 pages
- KON 509E: Design of Discrete Control Systems Introduction To Discrete Control Systems Assoc. Prof. Dr. Tufan KumbasarNo ratings yetKON 509E: Design of Discrete Control Systems Introduction To Discrete Control Systems Assoc. Prof. Dr. Tufan Kumbasar28 pages
- An Introduction To Control Theory Applications With Matlab100% (1)An Introduction To Control Theory Applications With Matlab152 pages
- Gujarat Technological University: W.E.F. AY 2018-19No ratings yetGujarat Technological University: W.E.F. AY 2018-195 pages
- Vel Tech High Tech DR - Ranagarajan DR - Sakunthala Engineering College - Department of ECENo ratings yetVel Tech High Tech DR - Ranagarajan DR - Sakunthala Engineering College - Department of ECE27 pages
- Feedback Control Theory: A Computer System's PerspectiveNo ratings yetFeedback Control Theory: A Computer System's Perspective52 pages
- Introduction To Discrete Control SystemNo ratings yetIntroduction To Discrete Control System34 pages
- McGraw-Hill - Digital Control Systems - Houpis & LamontMcGraw-Hill - Digital Control Systems - Houpis & Lamont
- Houpis, Constantine H, Digital Control Systems Theory, Hardware, SoftwareHoupis, Constantine H, Digital Control Systems Theory, Hardware, Software
- Chen ActiveNetworkAnalysis - FeedbackAmplifierTheory PDFChen ActiveNetworkAnalysis - FeedbackAmplifierTheory PDF
- Control Engineering - Theory, Worked Examples and ProblemsControl Engineering - Theory, Worked Examples and Problems
- Dr. Dong-Jin Lim - Control Systems Engineering - Design and Implementation Using Arm Cortex-M Microcontrollers-Independently Published (2021)Dr. Dong-Jin Lim - Control Systems Engineering - Design and Implementation Using Arm Cortex-M Microcontrollers-Independently Published (2021)
- 978-0134407623 Modern Control Systems (13th Edition)pdf download978-0134407623 Modern Control Systems (13th Edition)pdf download
- Full Download Modern Control Systems 13th Edition (Ebook PDFFull Download Modern Control Systems 13th Edition (Ebook PDF
- Download Complete Modern Control Systems 13th Edition (eBook PDF) PDF for All ChaptersDownload Complete Modern Control Systems 13th Edition (eBook PDF) PDF for All Chapters
- (eBook PDF) Modern Control Systems 13th Edition by Richard C. Dorf pdf download(eBook PDF) Modern Control Systems 13th Edition by Richard C. Dorf pdf download
- Control of Digital Systems: Automatic Control and System TheoryControl of Digital Systems: Automatic Control and System Theory
- KON 509E: Design of Discrete Control Systems Introduction To Discrete Control Systems Assoc. Prof. Dr. Tufan KumbasarKON 509E: Design of Discrete Control Systems Introduction To Discrete Control Systems Assoc. Prof. Dr. Tufan Kumbasar
- An Introduction To Control Theory Applications With MatlabAn Introduction To Control Theory Applications With Matlab
- Gujarat Technological University: W.E.F. AY 2018-19Gujarat Technological University: W.E.F. AY 2018-19
- Vel Tech High Tech DR - Ranagarajan DR - Sakunthala Engineering College - Department of ECEVel Tech High Tech DR - Ranagarajan DR - Sakunthala Engineering College - Department of ECE
- Feedback Control Theory: A Computer System's PerspectiveFeedback Control Theory: A Computer System's Perspective