Sie sind auf Seite 1von 47

Chapter 4

The Stability of Linear Control Systems
Outline
•Introduction,

•Routh‐Hurwitz Criterion,

•Routh‐Hurwitz Criterion: Special Cases,

•Steady‐State Error of Linear Control Systems

4/15/2018 By Awet Mana 2
Introduction 
• Three requirements enter into the design of a control system:
transient response, stability, and steady‐state errors.

• The most important problem in linear control systems concerns


stability. That is, under what conditions will a system become
unstable? If it is unstable, how should we stabilize the system?

• System stability can be defined with respect to the stability of the


natural response, or of the total response (natural + forced).
4/15/2018 By Awet Mana 3
Stability of the natural response:
oIf the natural response approaches 0 ⟶ ∞ , the system is stable. If 
the response approaches ∞ the system is unstable.
oIf the natural response remains constant or oscillates the system is 
marginally stable

Stability of the total response. Sometimes referred to as bounded‐input, 
bounded‐output (BIBO) stability:
oIf every bounded input yields a bounded output, the system is stable.
oIf any bounded input yields an unbounded output, the system is unstable
4/15/2018 By Awet Mana 4
• Considering the natural response, the system is stable as long as all closed‐loop system poles lie only in
the left half‐plane (LHP):

• Poles in the right half‐plane (RHP) yield exponential growth hence brings instability.
4/15/2018 By Awet Mana 5
• Pairs of poles on the imaginary axis lead to undamped sinusoids:
cos

• The presence of such poles make the natural response (at best) marginally stable.

• Poles of multiplicity greater than 1 on the imaginary axis lead to the sum of
responses of the form cos where n = 1; 2; . . . ; thus the amplitude
approaches infinity as time approaches infinity.

• Thus, unstable systems have closed‐loop transfer functions with at least one pole
in the right half‐plane and/or poles of multiplicity greater than 1 on the imaginary
axis.
4/15/2018 By Awet Mana 6
• Care must be taken here to distinguish between natural
responses growing without bound and a forced response, such as
a ramp or exponential increase, that also grows without bound.

• A system whose forced response approaches infinity is stable as


long as the natural response approaches zero.

4/15/2018 By Awet Mana 7
Using Matlab:
>>p=[1 3 2 3]
>>rootsp1=roots(p1)
rootsp1 =
‐2.6717 + 0.0000i
‐0.1642 + 1.0469i
‐0.1642 ‐ 1.0469i

• The oscillations for the stable system diminish. Also notice that the stable system’s response in
this case approaches a steady‐state value of unity.
4/15/2018 By Awet Mana 8
Using Matlab:
>>p2=[1 3 2 7]
>>rootsp1=roots(p1)
rootsp2 =
‐3.0867 + 0.0000i
0.0434 + 1.5053i
0.0434 ‐ 1.5053i

• The oscillations for the unstable system increase without bound.


4/15/2018 By Awet Mana 9
• System stability is determined entirely by the location of the roots of the
closed‐loop transfer function’s denominator. For the example above this
polynomial was, 3 2 , where k was either 3 or 7. 

• With k = 3 the system was stable, but becames unstable for k = 7.

• These roots are located wherever the following characteristic equation


holds,

3 2 0

4/15/2018 By Awet Mana 10
• If the denominator polynomial has either of the following characteristics, then it will
have RHP roots and the system will be unstable:
 If the polynomial coefficients differ in sign (i.e. not all positive or all negative). We have stability iff
the polynomial can be factored as follows where the real part of each root ri is positive:
···
This will always generate a polynomial with all positive coefficients (or all negative if we multiply 
everything by ‐1)
 If the polynomial has missing coefficients. If powers of s are missing, the system is either unstable or,
at best, marginally stable.

• One option is to determine the locations of the roots with Matlab. But we will need to 


establish analytical criteria for stability for a quick review of a system stability condition.

4/15/2018 By Awet Mana 11
Routh‐Hurwitz Criterion
• This criterion allows us to obtain a count of the number of poles in the LHP, RHP,
and on the .

• It does not tell us where the poles are, only how many there are of each category.

• We must first generate a Routh table. Assume our system is as follows (we care
only about the denominator from a reduced closed loop transfer function):

4/15/2018 By Awet Mana 12
Generating a Basic Routh Table
• Begin by labeling the rows with powers of s from the highest power of the
denominator of the closed‐loop transfer function to s0.
• Then start with the highest power and list every other coefficient in the first row of the
table. In the second row list the skipped coefficients in order.

4/15/2018 By Awet Mana 13
• We continue by filling in each entry as a negative determinant
divided by the first column of the row above.

• The left column of the determinant comes from the first entries of
the two rows above. The right column comes from the elements
of columns above and to the right.

4/15/2018 By Awet Mana 14
4/15/2018 By Awet Mana 15
• Example 1: Make the Routh table for the system shown in Figure below

4/15/2018 By Awet Mana 16
• Note: For convenience, we can multiply any row by a positive constant. This was done in the s2
row above.

4/15/2018 By Awet Mana 17
• Routh Table Interpretation: The number of RHP roots equals the number of sign
changes in the first column. (The proof of this statement is beyond the scope of this
course).

• For the example above there are two sign changes. Therefore the system has two RHP
poles and is unstable.

• Find the actual roots of the characteristics  of the equation using Matlab. 

>>p3=[1 31 10 1030] rootsp3 =
>>rootsp3=roots(p3) ‐31.7090 + 0.0000i
0.3545 + 5.6883i
0.3545 ‐ 5.6883i

4/15/2018 By Awet Mana 18
Practice problem 1

1. Make a Routh table and tell how many roots of the following polynomial 
are in the right half‐plane and in the left half‐plane.

2. Determine the stability of the closed-loop transfer function

4/15/2018 By Awet Mana 19
Routh‐Hurwitz Criterion: Special Cases
Zero in First Column

• If the first element of a row is zero, division by zero would be required to


form the next row.

• To avoid this phenomenon, an epsilon, ϵ, is assigned to replace the zero in


the first column.

• The value ϵ is then allowed to approach zero from either the positive or
the negative side, after which the signs of the entries in the first column
can be determined.
4/15/2018 By Awet Mana 20
Example 2
• Determine the stability of the closed‐loop transfer function

• Completed Routh table

4/15/2018 By Awet Mana 21
• Determining signs in first column of a Routh table with zero as first element in a row.

• Thus, there are two sign changes and therefore two RHP poles. The system is unstable.
By Awet Mana
4/15/2018 22
Entire Row is Zero 
• Sometimes while making a Routh table, we may find that an entire row
consists of zeros because there is an even polynomial that is a factor of the
original polynomial.
• This case is handled differently from the case of a zero in
only the first column of a row.
• Example 3: Determine the number of right‐half‐plane poles in the closed‐
loop transfer function

10
7 6 42 8 56
4/15/2018 By Awet Mana 23
• Routh table for  7 6 42 8 56

• Stop at the third row, since the entire row consists of zeros, and use the following procedure.
• Form an auxiliary polynomial, using the entries in the row immediately above the row of zeros
that row as coefficients. The auxiliary polynomial is, 6 8.

• Differentiate the polynomial with respect to s and obtain 4 12 0

• Use the coefficients of the derivative to replace the row of zeros.


4/15/2018 By Awet Mana 24
• Let us look further into the case that yields an entire row of zeros. An entire row
of zeros will appear in the Routh table when a purely even or purely odd
polynomial is a factor of the original polynomial.
Interpretation:
• We must split our interpretation and consider first the part above the ROZ
separately from the part below.
• Everything below the ROZ is a test of EP. In this case we see no sign changes
above or below. Therefore, there are no RHP poles. Is the system stable or
only marginally stable?

4/15/2018 By Awet Mana 25
• In the previous example the EP s4 + 6s2 + 8 was a factor of the original polynomial. EP’s have
the property that their roots are always symmetric about the origin.
• These roots can be purely real, purely imaginary, or complex, but each root will have a match
located 180o opposite.
• This means that a system with an even
polynomial factor will either be unstable
(A or C) or marginally stable (B).
• In our example s4 + 6s2 + 8 has four purely
imaginary roots (two pairs of roots as in B).
Therefore the system is marginally stable.

4/15/2018 By Awet Mana 26
Practice problem 2
1. For the transfer function 

tell how many poles are in the right half‐plane, in the left half‐plane, and on the jω‐axis. Use the Routh‐
Hurwitz criterion. Check out Example 6.3 of your text on Page 309.
2. Use the Routh‐Hurwitz criterion to find how many poles of the following closed‐loop system, T(s), are in the rhp,
in the lhp, and on the jω‐axis:

3. Find the range of gain, K, for the system of Figure below that will cause the system to be stable, unstable,
and marginally stable. Assume K > 0. Check out Example 6.9 of your text on Page 316.

4/15/2018 By Awet Mana 27
Steady‐State Error of Linear Control Systems
• Steady‐State Error. 

• Steady‐State Error for Closed‐Loop Systems. 

• Steady‐State Error for Unity Feedback Systems.
Static Error Constants and System Type.

• Steady‐State Error for Non‐unity Feedback Systems. 

4/15/2018 By Awet Mana 28
Steady‐State Error 
• Control systems are judged according to their performance in three areas:
Transient response, Stability, and Steady‐state error

• Steady‐state error is the difference between the input and the output for a
prescribed test input as → ∞ .
• Clearly we desire systems with zero steady‐state error. We will obtain the steady‐
state part of the total response simply by letting → ∞ and determining ∞ .
Hopefully it will tend to zero, or at least some small number.
• Important: A prerequisite to our analysis is that the system is stable. For an 
unstable system   ∞ ∞

4/15/2018 By Awet Mana 29
• Consider the total responses shown below for step and ramp inputs. 

• Steady-state error can occur because of nonlinearities in our system, but this is
beyond our scope. We consider error that arises because of the system itself
and its input.
4/15/2018 By Awet Mana 30
Steady‐State Error for Closed‐Loop Systems 
• Steady‐state error can be calculated from a system’s closed‐loop transfer
function, T(s),

• Steady‐State Error in Terms of T (s )



• But C(s)=T(s)R(s)
• E(s)=R(s)[1‐T(s)]
• This equation allows us to solve for e(t) 
at any time, t. 
General representation 

4/15/2018 By Awet Mana 31
• Given T(s) and a particular input R(s) we could take the ILT and let  → ∞ to obtain the 
steady‐state error  ∞ . However, there is a more direct way that also yields more 
insight...
• The Final Value Theorem 

e ∞ lim lim
→ →

∞ R(s)[1−T(s)]
→ →
• Example 4: What is the steady‐state error for R(s) = 1/s and T(s) = 5/(s2 + 7s + 10). 
• First we obtain E(s), 

• Applying the Final Value Theorem yields e(∞) = 1/2. 

4/15/2018 By Awet Mana 32
Steady‐State Error for Unity Feedback Systems 
• Consider the feedback control system shown in Figure below 

• We will establish E(s) in terms of the system and its input

1
• Apply the Final Value theorem: 
∞ lim ; Steady-State Error in Terms of G (s)
→ 1
4/15/2018 By Awet Mana 33
• We now determine ∞ for step, ramp, and parabolic input

4/15/2018 By Awet Mana 34
4/15/2018 By Awet Mana 35
4/15/2018 By Awet Mana 36
4/15/2018 By Awet Mana 37
4/15/2018 By Awet Mana 38
• Example 5: Find the steady-state errors for inputs of 5u(t), 5tu(t), and 5t2u(t) to the system
shown in Figure below. The function u(t) is the unit step

4/15/2018 By Awet Mana 39
• Example 6: Find the steady-state errors for inputs of 5u(t), 5tu(t), and 5t2u(t) to the system shown in Figure
below. The function u(t) is the unit step

4/15/2018 By Awet Mana 40
Static Error Constants
• The steady‐state error performance specifications are called static 
error constants. 
• The three terms in the denominator that are taken to the limit
determine the steady‐state error.
• We call these limits static error constants. Individually, their names are:
• position constant, Kp,  • The value of the steady‐state error decreases as
where the static error constant increases.
• An alternative method:

• Velocity constant, Kv,  • Thus, for a step input,


where
• For a ramp
input,
• Acceleration constant, 
Ka, where
• For a parabolic
input,
4/15/2018 By Awet Mana 41
System Type 
• Since steady‐state errors are dependent upon the number of integrations in the forward
path, we give a name to this system attribute.

• we define system type to be the value of n in the denominator or, equivalently, the
number of pure integrations in the forward path. Therefore, a system with n=0 is a Type
0 system. If n=1 or n=2, the corresponding system is a Type 1 or Type 2 system,
respectively.

4/15/2018 By Awet Mana 42
Example 7: For each system of Figure below, evaluate the static error constants and find the
expected error for the standard step, ramp, and parabolic inputs.

• For a step input,

• For a ramp input,

• For a parabolic input,

4/15/2018 By Awet Mana 43
• For a step input,

• For a ramp input,

• For a parabolic input,

4/15/2018 By Awet Mana 44
• For a step input,

• For a ramp input,

• For a parabolic input,

4/15/2018 By Awet Mana 45
• Relationships between input, system type, static error constants, and
steady-state errors

4/15/2018 By Awet Mana 46
Steady‐State Error for Non‐unity Feedback Systems 
• Often the feedback path has something other than unity gain. There may
be an output transducer and/or a compensating component in the
feedback path.

• Our approach is simply to transform such a system into an equivalent


unity feedback representation and then, use the previous methods
established. Either T(s) or G(s) method.

4/15/2018 By Awet Mana 47

Das könnte Ihnen auch gefallen