E E 380 - Linear Control Systems Supplementary Reading Transient Response Specifications For Underdamped Second-Order Systems
E E 380 - Linear Control Systems Supplementary Reading Transient Response Specifications For Underdamped Second-Order Systems
E E 380 - Linear Control Systems Supplementary Reading Transient Response Specifications For Underdamped Second-Order Systems
Supplementary Reading
Transient Response Specifications for Underdamped Second-Order Systems
The transient response of a practical control system often exhibits damped oscillations before reaching
steady state. Many of such systems can be modeled by underdamped second-order systems. In studying the
transient responses of such underdamped second-order systems, we typically consider the following transient
response performance specifications for their unit step responses: (1). Peak time Tp ; (2). Percent overshoot
P.O. (and maximum overshoot Mp ); (3). Rise time Tr ; (4). Settling time Ts . The definitions of these
specifications will be given below. Figure 1 shows an example of the unit step response of an underdamped
second-order system, with the performance specifications indicated on it. By using Matlab, we can zoom in
on the figure to determine the numerical values of these specifications.
2
Tp=1.603, P.O.=52.66%
1.4
1.2
y(t)
Tr=0.904
5% Ts=6.872
2% Ts=9.801
0.8
0.6
0.4
0.2
10
15
s2
n2
+ 2n s + n2
(1)
1 2 n t +
sin 1 2 n t), t 0
1 2
en t
y(t) = 1
sin( 1 2 n t + ), t 0
1 2
(2)
(3)
where = arccos . In the sequel, let us focus on the unit step response of this system and introduce each
of the above performance specifications along with their formulae.
Peak Time Tp : Peak time Tp is the time required for the response to reach the first peak of the overshoot.
To derive the expression of Tp in terms of and n , we first note from (2) that y(t) can be represented
as
y(t) = 1 et (cos d t +
sin d t), t 0
(4)
d
where = n is called attenuation and d = 1 2 n is called damped natural frequency. At Tp , since
y(t) reaches its maximum, we must have y (Tp ) = 0. Consider the equation y (t) = 0, we have
y (t)
= et (cos d t +
= (
2
+ d )et sin d t = 0.
d
(5)
T p = t1 =
=
.
(6)
d
1 2 n
Percentage Overshoot P.O. and Maximum Overshoot Mp : The maximum overshoot Mp is the
maximum peak value of the response curve measured from the final steady-state value, i.e.,
Mp = y(Tp ) y().
(7)
And the percentage overshoot P.O. is the percentage ratio of the maximum overshoot and the final steadystate value, i.e.,
y(Tp ) y()
100%.
(8)
P.O. =
y()
For the standard second-order system (1), substituting (6) into (4) we readily obtain
y(Tp ) = 1 e
(cos +
sin ) = 1 + e 12 .
d
(9)
Consequently we have
Mp = y(Tp ) y() = y(Tp ) 1 = e
and
P.O. =
1 2
y(Tp ) y()
100% = e 12 100%.
y()
(10)
(11)
Rise Time Tr : Rise time1 Tp is the time it takes for the response to rise from 0% to 100% of its final value.
To determine Tr , we only need to note that Tr is the minimum positive time instant such that y(t) = 1.
Using (3), the equation y(t) = 1 becomes
en t
sin( 1 2 n t + ) = 1.
1
2
1
The solutions to (12) are
solution gives us Tr , i.e.
k
d ,
(12)
arccos
=
.
d
1 2 n
(13)
We typically consider the 0% 100% rise time for underdamped second-order systems, although some texts also
consider the 10% 90% rise time too.
Settling Time Ts : Settling time Ts is the time required for the response curve to reach and stay within a
range about the final value of size specified by absolute percentage of the final value (usually 5% or 2%).
Instead of finding the actual Ts value, let us be content with an approximate value of Ts obtained as
follows.
Recall from our discussion of the peak value, the local maxima and minima of y(t) appear at tk = k
d ,
k = 0, 1, 2, . At these tk s, we have
|y(tk ) y()| = |y(
k
) 1| = etk .
d
(14)
Now instead of seeking the time required for the response curve to reach and stay within a range about the
final value, we want to determine the minimum k such that y(tk ) is within the range2. To determine such a
k value, we need to solve
X
(15)
|y(tk ) y()| = etk
100
where X = 5 or X = 2 (corresponding to 5% or 2% settling time). (15) leads to
tk ln(
or equivalently
X
)
100
(16)
k
X
).
ln(
d
100
(17)
d
X
ln(
).
(18)
100
We actually need to determine a minimum integer k which satisfies (18). However, to make the computation
d
X
ln( 100
) and substitute it into tk = k
easier to carry out, we roughly take k =
d to obtain the approximate
Ts as
X
)
ln( 100
T s tk =
.
(19)
T s tk =
5
)
ln( 100
3
3
=
.
(20)
2
)
ln( 100
4
4
=
.
(21)
The following table summarizes the performance specifications for the standard underdamped second2
order system H(s) = s2 +2nn s+2 discussed above.
n
Performance Specification
Peak Time Tp
Formula
Tp = d =
Maximum Overshoot Mp
Mp = y(Tp ) 1 = e
P.O. = e 12 100%
arccos
2
Tr =
d =
5% Settling Time Ts
2% Settling Time Ts
Ts
Ts
1 2 n
1 2
=
=
3
n
4
n
1 n
Note that since the |y(tk ) y()| is decreasing as k increases and each tk corresponds to a local maximum or
minimum of y(t), we must have y(t) to be within the range for every t tk as long as y(tk ) is within the range.
Remark. In the above table, all performance specification formulae except for Mp can be directly applied
to underdamped second-order systems with scaling factors other than 1. For Mp , the definition y(Tp ) y()
still applies; however, the term e 12 needs to multiplied by the scaling factor in order to compute Mp .
2
4
whose unit step response is
Example. Consider the underdamped second-order system H(s) = s2 +0.8s+4
shown in Figure 1. For this system, = 0.2 and n = 2. Applying the formulae in the table above, we can
compute the following performance measures.
= 1.603 sec
Peak time: Tp = d = 2 = 10.2
2 2
Maximum overshoot:
1 n
Mp = e 12
=e
0.2
10.22
= 0.5266
0.2
2 = arccos
= 0.904 sec
0% 100% rise time: Tr = arccos
10.22 2
1 2
5% settling time: Ts
2% settling time: Ts
3
n
4
n
=
=
1 n
3
0.22 =
4
0.22 =
7.5 sec
10 sec