You are on page 1of 12

Journal of Process Control 16 (2006) 91102

www.elsevier.com/locate/jprocont

Evaluation and simple tuning of PID controllers


with high-frequency robustness
Birgitta Kristiansson *, Bengt Lennartson
Department of Signals and Systems, Chalmers University of Technology, SE-412 96 Goteborg, Sweden
Received 18 March 2004; received in revised form 7 March 2005; accepted 3 May 2005

Abstract
An extensive study of robust and optimal tuning of PID controllers for stable non-oscillating plants is presented. It is built on a
set of well dened criteria related to output performance, stability margins and control activity. Dierent interesting properties of
the closed loop systems are observed. A set of simple tuning rules is based on these observations. These rules are compared to a
couple of well established tuning methods and are shown to give well competitive results, especially when simplicity, low control
activity and high-frequency robustness are emphasized. Derivative action is shown to improve performance signicantly compared
to PI control, with equal stability margin and a moderate increase of control activity, for most plants, including those with significant time delay.
2005 Elsevier Ltd. All rights reserved.
Keywords: PID control; Robustness; Optimization; Optimal systems; Time delay

1. Introduction
An objective method to evaluate the trade-o between performance and robustness in dierent frequency
regions for all kinds of linear controllers has recently
been presented by the authors [13]. This evaluation
method is briey presented in Section 2.
PI and PID controllers for a large amount of plant
models with all their poles on the negative real axis have
been optimized for good rejection of load disturbances
according to the suggested evaluation method. This is
motivated by the fact that most PID controllers work
as regulators [4,5]. For most investigated plants the optimal PID controller has a pair of complex zeros, which
has lead to a reformulation of the traditional PID con-

Corresponding author. Tel.: +46 31 7725739; fax: +46 31 7725731.


E-mail address: birgitta.kristiansson@chl.chalmers.se (B. Kristiansson).
0959-1524/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jprocont.2005.05.006

troller. The new formulation is presented in Section 3 together with the benchmark models used.
Section 4 is devoted to a series of general properties
for optimal PID controlled systems. It is noticed that
very limited improvements in terms of performance
can be attained, even with considerable raise of the control activity beyond a certain level. Moreover, the bandwidth of the optimal system is typically just above the
phase crossover frequency for the plant, and a PID controller designed for good disturbance rejection will in
most cases introduce a signicant negative phase shift.
It is also observed that the design method used will
place the resulting poles very similar for all plants
investigated.
An extremely simple method for design of PID controllers is presented in Section 5. All plant knowledge
that is asked for can be found by a simple step response.
An extended description of this and some related tuning
methods is given in [6,7]. The presented method is compared to some common and well appreciated tuning
methods. It is shown that the new method well asserts

92

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

itself. Compared to e.g. the appreciated IMC method,


which has just one parameter (k) to tune performance
against robustness, the presented tuning method also
highlights the possibility to optionally tune the control
activity and the high-frequency robustness. Derivative
action may then be introduced without too large
sensitivity to sensor noise, which used to be the reason
why derivative action is often excluded in process
industry.
It has sometimes been stated that derivative action
should not be recommended for plants with signicant
time delay [5,8]. In Section 6 is shown that derivative action, also for these plants, will mostly improve the output performance of the closed loop system remarkably
compared to a PI controller. The price is a moderate increase of the high-frequency gain. However, it is important that all accessible parameters in the PID controller,
including a low-pass lter on the derivative part, are included in the design procedure.

2.1. Performance criterion

2. Evaluation method

With robust design and small undershoot in the load


disturbance step response, IAE is almost equal to the
integral of the error IE and according to [9]
Z 1
1
IE
et dt
ki
0

The proposed performance criterion is mainly a measure of the systems ability to handle low-frequency load
disturbances, a frequency domain alternative to the
more commonly used criteria based on a function of
the error signal [9,10]. It is formulated as




1 
1




J v  Gyv   S v s
1

s
s
1
1
When there is integral action in the controller, L(s) =
G(s)K(s)  1 and Sv(s) = G(s)/(1 + G(s)K(s))  K1(s) 
s/ki for low-frequencies (ki is the integral gain). Thus, in
this frequency range Jv  1/ki has the advantage of
being almost independent of the plant model.
Alternative performance criteria: There are several different performance criteria proposed in the literature.
The most commonly used is presumably
Z 1
IAE
jetj dt
0

Consider the SISO system in Fig. 1, where a plant


G(s) is controlled by a controller K(s). It has three inputs: the reference signal r(t), the process disturbance
v(t) and the sensor noise w(t). Relevant outputs are the
controlled output y(t), the control signal u(t) and the
control error e(t) = r(t)  y(t). Also introduce the loop
transfer function L(s) = G(s)K(s) and the following four
sensitivity functions with corresponding closed loop
transfer functions, which have related output and input
signals as indices.
Sensitivity function Ss

1
Ger s
1 Ls

Complementary sensitivity function


Ls
Gyr s Gyw s
T s
1 Ls
Disturbance sensitivity function

S v s

Gs
Gyv s
1 Ls

As is illustrated in Fig. 2, and also noted above, there


is a close relation between Jv and ki, and thus between Jv
and IE/IAE. The integral gain ki is equal to the intersection with the frequency axis of the low-frequency
asymptote to jSvj, while 1/Jv is the intersection for the
tangent of jSvj in the same direction. Typically the distance between those two points on the frequency axis
is small. However, Jv as a criterion has the advantage
above ki of not only measuring the low-frequency
asymptote, but also Sv somewhat higher up in the lowto mid-frequency region. Moreover, Jv is straightforward to use as a performance criterion even for MIMO
systems. For more details about Jv and ki, see [2].

Control sensitivity function


S u s

Ks
Gur s Guw s
1 Ls

10

ki

1/J

|Sv|

v
r

+
-

K(s)

G(s)

Fig. 1. Closed loop SISO system with plant G(s) and controller K(s).

10

-1

10

Fig. 2. The gure shows the relationship between ki and Jv for a


system with a typical Sv gain.

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

The criteria discussed so far are all mainly related to


the low-frequency region, but there are also interesting
performance aspects related to the mid-frequency range
e.g. bandwidth xb and settling time ts. Fig. 3 shows the
reference and load disturbance step responses for a PID
controlled plant, with the closed loop system optimized
according to dierent criteria. It is obvious that Jv has
the potential to reward good process disturbance compensation; actually Jv together with IAE is the criterion
that brings about the best transient behavior for load
disturbances. To improve the reference response, J e
k 1s Gry k1 k 1s Sk1 can be used as an alternative performance criterion. Moreover, it is always possible to optimize the closed loop with respect to Jv and then design a
reference signal prelter to get the demanded reference
response.

93

A suitable mid-frequency robustness criterion, the


generalized maximum sensitivity GMS can then be formulated as
GMS maxkSk1 ; akT k1

This criterion converts the measure (3) to a corresponding MS level. A similar criterion was formulated
by Schei [19]. In [9] two suitable values for MS (1.4
and 2.0) are proposed. Throughout this paper the default values are MS = 1.7 and MT = 1.3, which means
that an optimal controller with GMS = 1.7 results in a
system with Gm P 2.4 and um P 45. For high-gain
controllers, such as Smith predictors, a demand on
bounded high-frequency loop-gain may also be included
in GMS, for more details see [2].
2.3. Control activity

2.2. Mid-frequency robustness


Two classic measures are still common to characterize
the mid-frequency MF robustness, the phase margin um
and the gain margin Gm [1115]. However, in recent
years a restriction of the maximum sensitivity function
kSk1 max jSjxj 6 M S

has been more and more accepted as an exclusive


robustness measure [9,1618]. The reason is that kSk1
is equal to the inverse of the minimum distance from
the loop transfer function to the critical point (1, 0)
in the Nyquist plot. In many situations it is also a fully
sucient MF robustness measure.
With demands on further damping of the step response or increased phase margin without slowing down
the system response, adding a restriction on the maximum complementary sensitivity function can be a good
alternative
kT k1 6 M T

especially for unstable plants including those with integral action, see [19,20,8,21].

1.2

This is a mid- to high-frequency MHF criterion, since


Guw typically has its maximum around or slightly above
the closed loop bandwidth. In addition to the criteria Jv,
GMS and Ju, a fourth criterion for the HF region, a
measure of robustness against HF model uncertainties
and sensitivity to HF sensor noise, can be introduced.
This is motivated in cases with roll-o in the controller,
see [1,2].
2.4. Evaluation procedure
To design a controller is to modify a set of tuning
parameters q. A reliable evaluation method is to keep
two of the above introduced criteria xed or at least restricted and then tune q to the minimum of the third criterion. For example, evaluation of LF performance is

max ki

0.2

1
0.8

J u kGur k1 kGuw k1 kS u sk1

0.3

min J

Design of a control system is typically a trade-o between performance and its price in terms of control
activity, as soon as demanded stability margin is guaranteed. Therefore, introduce the control activity criterion

min t

min ts

min IAE

0.1

min IAE

0.6
max b

0.4

min Jv

0.2

max

max k

0
0

10

0.1

Fig. 3. Reference (left) and load disturbance (right) step responses for G(s) = e

0.3s

10

/(1 + s) with a PID controller optimized for dierent criteria.

94

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

accomplished by solving the constrained optimization


problem
minq J v q GMS q 6 C 1

J u q 6 C 2

where the constants C1 and C2 may be given dierent


values. The default value of C1 in this paper is 1.7, while
the value of C2 may vary.
By this optimization procedure completely dierent
controllers may be compared under equal conditions.
In fact it is straightforward to include even sampleddata controllers with dierent sampling periods in such
a comparison [22]. An evaluation method like this one
is asked for in [23]. A similar idea, but with other criteria
and with more vague constraints, is presented in [24] and
another one in [25].
The expression optimal controller is from now on used
for a controller which is optimized according to (6) with
all available controller parameters included in the tuning
vector q. In this paper Matlab Optimization Toolbox is
used for the computation.

3. The PID controller and the plant models


The traditional PID controller with a low-pass lter
on the derivative part used to be formulated as


1
sT d
K PID s K p 1

7
sT i 1 sT f
Very often, ever since the days of Ziegler and Nichols,
the relation a = Ti/Td has been given a xed value a = 4
[2628], corresponding to a double zero in a controller
without derivative lter. In the same way b = Td/Tf
has been xed, often to 10, [2931].
3.1. Alternative tuning parameters
In e.g. [32,21] it has been shown that a PID controller, optimized with the lter included and all parameters
free, often implies complex zeros in the controller. This
has been veried for a large number of plants with poles
strictly on the negative real axis. Hence, a suitable alternative parameterization of the PID controller is
K PID s k i

1 2fss ss
s1 ss=b

Tf

s
b

T i 2fs  T f

Td

s2
 Tf
Ti

K p K iT i

3.2. Plant knowledge


To be accepted in industrial applications, controller
tuning rules must be based on a limited amount of plant
knowledge that is comfortable to nd.
When the intersection of the Nyquist plot of the
plant with the negative real axis is known, this knowledge together with the static gain can form a sucient
base for formulation of tuning rules. The plant can then
be characterized by its j value, j jGjx180G j=G0, a
number proposed in [33] as a measure of the complexity
of the plant. Here x180G is the frequency where the
plant has a phase lag of 180, which corresponds to
the intersection point. This j number usually falls in
the interval [0, 1]. Sets of tuning rules based on this jnumber have been published by the authors in [1,32,
21,7,3]. Over the time they have been improved and
simplied.
However, when a relay experiment including hysteresis according to [27] is used to nd the plant knowledge,
the controlled self-oscillating frequency will not be
x180G , but somewhat lower, typically around x150G . Then
j may be modied to
j150

jGjx150G j
G0

10

This character is also useful for cases when x180G is


too high or does not even exist. A set of tuning rules
based on j150 is given in [7,2]. The relation between
the two j-numbers is j150  0.9j + 0.16 when 0.1 6
j < 0.9.
An attractive alternative to the frequency character j
or j150 is a time domain character. In [6] it is shown that
T63, the time it takes for a step response to reach 63% of
its nal value, is a successful candidate.

with the four tuning parameters the integral gain ki, the
high-frequency gain k1 = KPID(1) = kisb, the zero
damping f and the zero time constant s. The notation b
is short for k1/(kis), but may also be used as a design
parameter, an alternative to k1.
The PI controller may then be regarded as a PID controller with f = b = 1 and may be written as
K PI s

When it comes to implementation it is straightforward to translate these new design parameters to the traditional ones in (7).

ki
1 ss
k1 ki
s
s

3.3. Plant models


The models used for the investigations in this paper
are approximately the same as in [3]. These models are
essentially those recommended in groups 15 in [34],
suggested as standard benchmark models for testing of
PID controllers. They include all kinds of plants with
their poles strictly on the negative real axis. There are
plants with time delay or non-minimum phase zeros,
plants of high and low orders, plants with multiple
and spread poles, etc.

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

p
p
4  8 2a2 a for k 1 4a 1  2 2a2 a=Gm K

4. Optimal system properties


Some general system properties can be observed
when PID control is applied and all parameters including the lter constant are used to optimize the closed
loop system according to (6).
4.1. The relation between Ju and k1
According to (5), the control activity criterion Ju is
equal to the maximum of the control sensitivity function
Su(x). For high frequencies, where jS(x)j  1, it can be
observed that jSu(x)j  jK(x)j  k1/xm = K(1)/xm.
For the PID controller (8) m = 0 and, due to the derivative action, maximum of Su(x) typically occurs when
x ! 1, meaning that Ju = k1.
In the PI case the maximum of Su(x) occurs for lower
frequencies, most often just above the closed loop bandwidth. However, also for the PI controller there is a
close relation between Ju and k1. Higher values of Ju
correspond to higher values of k1, at least for plants
with not too high j values, see [21]. Fig. 4 illustrates
the relations between Ju and k1.
4.2. The trade-o between Jv and Ju
In Section 2 we noted that Jv  1/ki and just above we
found that there is a close relation between Ju and k1.
Consider a rst-order delayed plant, where the delay
is approximated by a rst-order Pade approximation
Gs

95

K
K1  0.5sLd
esLd 
1 sT
1 sT 1 0.5Ld

When this plant is controlled by a PI controller (9), and


a gain margin Gm is demanded, straightforward calculations using RouthHurwitz criterion show that 1/ki =
f(k1) has a minimum 1=k i min Ld Gm K=12a

where a = T/Ld. Extensive empirical investigations, presented in e.g. [21] show that there is most often a corresponding minimum in the Jv/Ju-graph, when the
controller is a PI one. This fact is also observed in [35]
and illustrated by Fig. 5 (see also [3]).
In the optimal PID case there is normally no such
minimum. On the contrary, the Jv-graph tends to approach a horizontal asymptote as Ju grows, cf. Fig. 5.
For a non-minimum phase system it can even be proved
theoretically that a lower limit exists, see [2]. Thus it can
be argued that it is useless to increase Ju above a certain
level J uec , because the reward in terms of decreased Jv is
too small or none. This tendency towards non-decreasing Jv is more obvious and the economic level J uec is
lower the more complex the plant is (higher j value),
see [3]. The economic limit is obviously not very
sharp, but for the plant in Fig. 5 it has been estimated
to approximately 10.
It should be emphasized that an optimal PID controller, working at its economic level, mostly oers better
system properties than an optimal PI controller, and still
J uec is very reasonable. Moreover, it is seen in Fig. 5 that
when a PID controller with a = 4 [2628] and b = 10
[2931] oers the same performance Jv as the optimal
PI controller, the demanded control activity Ju is almost
5 times higher. Since these values of a and b are often
recommended and used in practice, this is one obvious
reason why derivative action is not used in most industrial applications. In some (especially older) process control systems, the parameter b is xed to b = 10 and can
not even be adjusted.
It is also worth noting that for those Ju values, where
the PI controller works at its best, the PID graphs typically come close to the PI graphs. This is what could
be expected, since then b ! 1, and a PID controller with
b = 1, f = 1 corresponds to a PI controller.

20
v

k PID
10

PID optimal
PID:a=4,b=10
PID:a=4,b opt

|Su|PID

PI optimal

Ju
2

|S |

u PI

PI

1
0.5 2
10

10
0.3s

10

Fig. 4. Plant G = e
/(1 + s) optimally controlled by a PI controller
(dashdotted), a PID controller (solid). The gure illustrates the
relations between k1 and Ju.

12

Fig. 5. Jv as a function of Ju for the plant G(s) = e0.3s/(1 + s)2


controlled by dierent PI(D) controllers. In all cases GMS = 1.7, while
Jv is minimized for each Ju value, corresponding to dierent C2 in (6).

96

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

4.3. The PID controller zeros

2.5
GM S

The PID controller zeros are in (8) characterized by


two parameters, the damping ratio f and the time constant s.
For the optimal f no signicant dependence has been
found neither on j150 nor on Ju for stable non-oscillating
plants [2,6]. Fig. 6(a) shows that f = 0.75 is in most cases
a good approximation. That corresponds to a = Ti/Td 
(2f)2 = 2.25, signicantly less than the standard value 4.
Fixing f to this value and optimizing the remaining
parameters will always give results that are almost
impossible to separate from the optimal ones.
Also the optimal natural frequency 1/s is almost independent of Ju, but normalized by x150G , it increases almost linearly with j150 see Fig. 6(b).

compared to what is reached with standard values such


as a = 4 and b = 10. This is illustrated by Fig. 5.

4.4. The lter factor

4.5. Optimal pole placement

Among industrial users there is a widely spread scepticism against derivative controller action, based on anxiety about gain of sensor noise. This has caused the
derivative function in many controllers to be switched
o. When derivative action, despite this, is used, a
low-pass lter is always included in the control loop to
bound the HF gain. However, tuning of the controller,
e.g. for demanded stability margins, is mostly derived
without consideration of the lter, which is added afterwards. The lter factor is then typically given an ad hoc
value, low enough not to inuence the tuning. Fig. 7 offers a hint about why b =pTd/Tf P 10 has often been
chosen. Note that b  b= a, why a = 4 and b = 10
approximately corresponds to b = 20. For Fig. 7 the
parameters s, f and ki have been kept constant, while
b has been changed and GMS observed for G(s) =
e0.3s/(1 + s)2. It is obvious that b has a signicant
impact on the stability margin, but that this impact
decreases for b > 20. When the lter factor is included
in the tuning parameter vector q, the properties of a system can be remarkably improved (low Jv and low Ju)

When PID controllers have been optimized for a


great amount of dierent, stable and non-oscillating
plants without zeros in the negative half-plane, striking
regularities have been noticed regarding the closed loop
poles. For plants with j150 in the interval 0.10.5 the
slowest pair of poles have a damping factor very close
to 0.6 and a relative natural frequency (x0/x150G) close
to 0.65, see Fig. 8. The next pair of poles have a damping close to 0.5 and a relative natural frequency close to
1.4. The rest of the poles may be real or complex, but
their real parts are at least 6.5 times the real part of
the second slowest pair of poles. When they are complex
they come close to the real axis (damping factor close to
one).

1.5
2

10

20

100

500

Fig. 7. Inuence on GMS of changes in b for G(s) = e0.3s/(1 + s)2.

4.6. Bandwidth and crossover frequency


It may be argued that the closed loop bandwidth xb,
obtained with the proposed evaluation method, is often
rather small. The same assertion is valid for the open
loop gain crossover frequency xc. This is a consequence

1/( 150G)

1.5

0.5

0.5

0
0
a

0.5

150

0
0

1
b

0.5

150

Fig. 6. Zero parameters, damping f and normalized natural frequency 1/s, as functions of j150, optimal outcomes (+) and linear approximations.

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

cl

97

1.5

0/150

1:st pole pair

0.5

2nd pole pair

2:nd pole pair

0.5
1:st pole pair

0
0

0.2

0.4

150

0.6

0
0

0.2

0.4

150

0.6

Fig. 8. Dominating closed loop pole pairs as functions of j150 for optimal PID controllers with GMS = 1.7 and J u  J uec .

of the minimization of Jv  1/ki, the relatively strong


demands on MF robustness (stability margin) and the
moderate control activity. According to Fig. 9 the crossover frequency decreases slightly with increasing j,
xc  0.6  0.35jx180G , while the resulting closed loop
bandwidth is almost independent of this value, xb 
1.1x180G (x180G plant phase crossover frequency). This
means that xb  23 times xc, which is signicantly larger than the rule xc 6 xb 6 2xc, found in the literature
[36].

Sometimes it has been claimed that when a phase margin of 45 (approximately corresponding to MT = 1.3) is
demanded, a well-tuned PID controller conserves the
phase shift of a minimum phase plant at a frequency
where it is around 135 [37,5]. This means that the controller at the open loop gain crossover frequency xc will
have no signicant phase lag. According to [37] it is even
likely to have a positive phase at this frequency.
Fig. 10 shows the phase shift Du found for optimal
PID controllers at xc for dierent values of j. It indi-

1.5

0
0

30

60
0

0.5

cates that Du at this frequency increases negatively with


j, roughly as Du = 5100j. This result is valid for all
kinds of stable non-oscillating plants investigated in this
paper, including minimum phase plants with higher j
values. The optimization means a minimization of Jv,
which approximately corresponds to a maximization
of the integral gain ki. It is well known that large integral
action brings good disturbance compensation but also
large negative phase shifts.

5. Tuning rules for PID controllers


Based on observations similar to those in Section 4,
tuning rules for PID controllers have been formulated,
see [3,2,6]. Here only two sets will be reproduced, one
based on j150, and one extremely simple rule based on
T63. These rules are then compared to a couple of well
established tuning methods.

b/180

0.5

PID

Fig. 10. Controller phase shift (DuPID) at xc as a function of j for


optimal PID controllers with GMS = 1.7 and Ju = 8.0.

4.7. Controller phase shift

c/180

5.1. Simple almost optimal tuning rules


0.5

Fig. 9. Normalized closed loop bandwidth xb and gain crossover


frequency xc as functions of j for optimal PID controllers with
GMS = 1.7 and Ju = 8.0.

For tuning of PID controllers for plants with all their


poles strictly on the negative real axis, the following
rules can be recommended.

98

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

1
x150G 0.44 0.86j150
s


1
2
min 3
k1
; 25
G0
j150G


x150G
0.45
ki
 0.1
G0 j150 0.07

5.2. Comparison between dierent tuning rules

f 0.75

11

Tuned by these rules, the closed loop systems will show


stability margins GMS in the interval 1.651.85 for almost all plants tested (j150 = 0.030.9), in most cases
close to 1.7. The resulting Jv value diers from the optimal one at corresponding Ju in most cases with less than
5%. The only exceptions noted are plants with very long
time delays (Ld/(T63Ld) > 4 corresponding to j150 P
0.95). The rules for f and s are valid over a wide interval
of Ju-values. This implies that the user can apply them
and then make his own trade-o between control activity (k1) and output performance (ki) as well as stability
margins (GMS).
For formulation of still more simple tuning rules the
parameter k1 can be exchanged by b, which for the
most common plants may be given a value around 8.
If the plant has very simple or a little more complex
dynamics, b may be slightly decreased or increased, see
[6]. Then an extremely simple tuning rule can be
formulated.
From a step response giving T63, set f = 0.8, s = T63/3
and b = 8.1 Tune the only remaining parameter ki to demanded stability margin (damping).
The results of this simplied rule for plants with
0.09 6 j150 6 0.9, including e.g. plants with highly
non-minimum phase behavior, are remarkably close to
the optimal ones. For plants with j150 < 0.09, and plants
with rst-order dynamics plus time delay, a lower s
value, and in the second case a somewhat higher f value,
are recommended, see further details in [6].
Using this method, it cannot be argued any more
that a PID controller is more dicult to design than
a PI controller. PI control, where b = 1 and f = 1,
means that the control activity Ju is determined by
the optimal solution in the Jv/Ju curve, cf. Fig. 5. Most
often this Ju-level may be increased somewhat. When
derivative action is introduced with a b value of 510,
performance is normally signicantly improved, without serious increase of sensitivity to sensor noise and
high-frequency model uncertainties. The suggested
method highlights the possible trade-o between this
noise sensitivity, the ability to compensate load disturbances and stability margins, by sequentially tuning b
and ki.

Be prepared to reduce it somewhat if the high-frequency gain is


considered too high, possibly down to b = 5 for plants with simple
dynamics.

Results in terms of Jv, Ju and GMS for three dierent plants, from eight common tuning methods are
compared with each other and with the optimal controller (according to (6) with J u  J uec ), see Tables 13. For
all controllers the gain is adjusted to achieve GMS = 1.7,
original values given in parentheses. When no special lter is proposed, a rst-order low-pass lter with b = 10 is
added to make the controller proper.
ZieglerNichols (ZN): ZieglerNichols method
from 1942 [26] is included because it is classical and
well-known, also among practicians and students.
H)
AstromHagglund (AH): The method named (A
is introduced in [9]. It is an empirical frequency response
method, based on dominant pole design [38]. Two demands on MS (1.4 and 2.0) are proposed. Here the mean
values of the parameters given by the two demands are
used.
Internal model control (IMC): The IMC method is
introduced in [39,40]. Since a second-order plant model
is required for the method to result in a PID controller,
a model reduction according to [41] has preceded the
parameter calculation when necessary.
The IMC controller cancels the stable plant poles by
the controller zeros, which determines f and s. The
closed loop poles are placed as a double pole (IMC)
or alternatively (IMCZ), for plants with a non-minimum
phase zero at s = z, one pole in the stable image at
s = z. The characteristic polynomials of the closed
loop poles are then P(s) = (s + 1/k)2(IMC) or P(s) =
(s + 1/k)(s + z) (IMCZ), where k is the remaining tuning
parameter.
Skogestad IMC (SIMC): This method is a
modication of IMC introduced in [8]. The included
model reduction guarantees real poles. Since b is theoretically innite, only one nite pole remains to be
placed by the characteristic polynomial P(s) = (s + 1/
k). For fast response k is recommended to equal the delay time of the reduced model, and then there is no
parameter left for adjustment. In [8] b  100 is used in
the analysis, which results in large Ju values. In the nal
implementation a smaller value is recommended.
Hence, for SIMC2 the lter factor b is adjusted to give
J u J uec .
Traditional parameter choice (ab): In Section 3 attention was called to the fact that a = Ti/Td and b = Td/Tf
in the literature very often are given the xed values 4
and 10, respectively. For this study these values are used
and Ju is bounded by J uec .
KristianssonLennartson (KLj150): This tuning
method is presented above in (11). It is based on j150
and x150G .
KristianssonLennartson (KLT63): This extremely
simple method is based on s = T63/3, f = 0.8, b = 10,
and ki tuned to fulll GMS = 1.7.

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

99

Table 1
Comparison of dierent methods for design of a PID controllers (the plant model is G = 1/(1 + s)4)
Method

GMS

Ju

Jv

Optimal
ZN
H
A
IMCk=1.1
SIMC
SIMC2
ab
KLj150
KLT63

1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70

8.06
15.64 (24.1)
17.8 (17.8)
2.99
74.3 (49.6)
8.06
8.06
8.06
9.07

1.62
2.01
1.82
2.16
2.00
2.16
2.76
1.64
1.71

(2.32)
(1.70)
(1.44)

(1.75)

IAE
(1.30)
(1.81)
(3.00)

(1.57)

1.93
2.01
1.83
2.32
2.35
2.55
3.49
1.90
2.10

(1.40)
(1.82)
(3.13)

(1.82)

7.05
20.0
20.4
3.46
122
14.3
21.0
7.57 (7.25)
10.0

0.70
1.03
0.95
0.80
1.03
1.03
0.98
0.75
0.80

1.65
1.57
1.59
1.87
1.22
1.22
0.97
1.67
1.45

Table 2
Comparison of dierent methods for design of a PID controllers (the plant model is G = 1/((1 + s)(1 + 0.7s)(1.0.72s)(1 + 0.73s)))
Method

GMS

Ju

Jv

IAE

Optimal
ZN
H
A
IMCk=0.66
SIMC
SIMC2
ab
KLj150
KLT63

1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70

8.67
17.7 (28.0)
20.9 (21.5)
3.40
108.4 (85.0)
8. 67
8.67
8.67
11.1

0.88
1.04
0.95
1.33
0.92
1.11
1.63
0.89
0.88

1.05
1.05
0.96
1.35
0.92
1.11
2.06
1.03
1.07

6.75
20.0
20.7
3.69
102.9
9.89
21.0
7.04 (6.68)
10.0

0.71
1.03
0.96
0.87
1.00
1.00
0.98
0.75
0.80

1.00
0.92
0.96
1.22
0.97
0.97
0.60
1.05
0.91

13.1
20.0
24.3
4.29
3.99
100
15.2
21.0
13.2 (11.5)
10.0

0.89
1.03
1.12
1.00
1.00
1.00
1.00
0.98
0.75
0.80

0.88
0.91
0.73
1.00
1.00
1.00
1.00
0.67
0.85
0.97

(2.32)
(1.73)
(1.52)

(1.77)

(0.66)
(0.93)
(1.18)

(0.86)

(0.72)
(0.93)
(1.18)

(0.98)

Table 3
Comparison of dierent methods for design of a PID controllers (the plant model is G = (1s)/(1 + s)2)
Method

GMS

Ju

Jv

Optimal
ZN
H
A
IMCk=0.61
IMCZk=0.81
SIMC
SIMC2
ab
KLj150
KLT63

1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70
1.70

6.11
8.21 (12.0)
7.64 (7.23)
1.65
1.52
41.0 (50.0)
6.11
6.11
6.11
4.52

1.98
2.21
2.32
2.61
2.62
2.44
2.48
2.49
2.26
2.23

(2.39)
(1.65)

(2.01)

(1.86)

IAE
(1.51)
(2.45)

(2.00)

(2.12)

Results and comments: In Tables 13 the evaluation


criteria introduced in Section 2 in this paper, GMS, Ju
and Jv are noted and, moreover, IAE is added. The tuning parameters corresponding to (8), b, f and s are given. The fourth one ki is very close to 1/Jv.
That tuning by ZN mostly results in small Jv but
large (poor) GMS and large Ju is well known and con H in most cases implies competitive Jv but
rmed. A
to the price of relatively high Ju.
In IMC there is only one parameter available for the
trade-o between performance, damping and high-frequency gain. The demanded double pole leads to extremely low Ju values, clearly lower than J uec , but bad Jv
and IAE values. The model reduction according to
[41] typically obtains slightly complex poles, which is

2.59
2.63
2.80
3.03
3.04
2.84
2.90
3.20
2.98
2.87

(1.98)
(2.91)

(2.42)

(2.81)

favorable, since the optimal damping of the controller


zeros usually is close to f = 0.8. Remind that IMC is often applied in industrial applications. SIMC, on the contrary, results in very high Ju values, when b  100 is
used. The alternative SIMC2 with b adjusted to
give J u J uec is therefore included. The comparison
conrms that pole cancellation, which is part of both
IMC and SIMC methods, seldom results in optimal performance. The pole placements given by IMC and
SIMC may be compared with the optimal ones mentioned in Section 4.
KLj150 is based on J u  J uec . In KLT63 Ju is given
by b. In both cases Ju is kept on a reasonably low level.
This level is in general free to be chosen, while ki is left
for adjustment to a suitable damping.

100

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

J u 10  J uec . The step responses show in all cases signicantly smaller integrated errors for the PID controllers than for the PI controllers. The plants with delay all
through show better trade-o between Jv and Ju and
more favorable step responses than the corresponding
lag plants, despite the same j, GMS and Ju. However,
the prot oered by the derivative action is almost the
same for the two kinds of plant dynamics.
The reason, why derivative action is argued to be
more protable for plants with lag than for time delayed
plants, is possibly that an arbitrary linear controller,
with very large control activity Ju, theoretically can reduce Jv to zero for plants with lag. This is not the case
for time delayed plants, due to their non-minimum
phase behavior, see [7]. However, with reasonably large
control activity, the dierences between PI and PID control are comparable for plants with equal j value, independently of whether the plant dynamics include a
signicant time delay or not. Very large time delays,
on the other side, decrease the prot of derivative action, as is stated below.

6. PI or PID control for time delayed plants?


In the recent literature dierent opinions can be
found on whether derivative action should be used or
not, when there is a signicant time delay in the plant
[42,5,4,8,43].
6.1. Moderate time delays
For two plants of dierent types the benets from
derivative action are compared. Their transfer functions
are
1
1 s1 as1 a2 s1 a3 s
a 0.3; 0.4; . . . ; 1.0

Glag s

12

sLd

Gdelay s

e
1 s1 0.2s

13

For each a in Glag, Ld in Gdelay is varied to give the same


j value for the two plants. Fig. 11 shows the Jv/Ju relations and the process disturbance step responses for two
pairs of plants with equal j values, controlled with and
without derivative action. The Jv/Ju relations show
that, for both types of plants, Jv will decrease to
approximately half of its value for the optimal PI-controller, when derivative action is introduced with

6.2. Large time delays


For a pure dead-time process derivative action can be
shown to give no benet, see [42,5,43,44]. For plants

1.2
PIlag

PI

0.6

=1

PID

0.4
Jv 0.5

PIdelay

PIDlag

PI

0.5

0.2

PID0.5

PIDdelay

0
0.2

10
Ju

15

20

10

3.5

20

b
0.6

PIlag

PI

L =0.4
d

0.4
PID

lag

Jv

PID

0.4

y
PI

0.2

0.2

PID

0.2

PIdelay

PIDdelay

0.5
0
c

10
Ju

15

20

0
d

6
t

Fig. 11. Left: Jv/Ju relations for Glag and Gdelay with (a) j = 0.147 (a = 0.5,Ld = 0.211) and (c) j = 0.249 (a = 1.0,Ld = 0.405). Right: Process
disturbance step responses for corresponding plants (b) Glag + PID (Ju = 10) and Glag + PI (Ju optimal), (d) Gdelay + PID (Ju = 10) and Gdelay + PI
(Ju optimal).

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102


Table 4
Comparison of performance with PI and PID control
Plant model
1.5s

e
/(1 + s)
e5s/(1 + s)3
1/(1 + s)4
1/(1 + s)8

s2

Ld

J vPI (opt)

J vPID J uec

J vPID =J vPI

1
1.5
1.5
1.5

1.5
5.5
1.5
5.5

3.7
10.6
3.4
9.2

2.3
7.5
1.6
5.8

0.64
0.71
0.47
0.64

with long delay times, but also lag, there are dierent
opinions. According to [42] derivative action should be
used in those cases, while [5] means that derivative action should always be avoided for plants with signicant
time delays. In [8] a more precise recommendation says
that derivative action should only be used, when the second largest process time constant is greater than the delay time after model reduction (according to a given
scheme).
Table 4 shows a comparison of performance for optimal PI and PID controllers for some plants, where
derivative action according to [8] is not recommended
(s2 is the second largest process time constant according
to [8]). It is obvious that for all plants in the table derivative action has a contribution to oer. More general
however, there is no doubt that the improvement in performance oered by derivative action decreases when
Ld/Teq increases (Teq = eective time constant). When
Ld/Teq = 10 the improvement is approximately 15%
and for Ld/Teq = 20 it is 2% This is quite in accordance
with related results presented in [43,44].

7. Conclusions
In this paper a series of observations about closed
loop systems with optimal PI and PID controllers is presented. The expression optimal refers to controllers
that have the best possible trade-o between output performance and control activity when demanded stability
margin is guaranteed. The applied evaluation method
is briey presented in the beginning of the paper.
It has been noticed that the closed loop bandwidth xb
is very close to the phase crossover frequency of the
plant x180G and signicantly higher than the loop gain
crossover frequency xc. Furthermore, xc decreases with
j, while xb is independent. The pole placement normalized by the plant time scale is almost the same for most
optimal systems.
Observations that motivate alternative tuning parameters for the PID controller are reported. Based on
these observations simple, but still almost optimal,
tuning rules are presented. Since a PID controller
undoubtedly can oer signicantly improved performance compared to a PI controller, and the introduced
tuning rules are extremely simple to use, although they
include tuning of the low-pass lter in the derivative

101

part, they remove all reasons to avoid derivative action


in industrial applications. The introduced rules imply
economically eective low control activity, corresponding to reasonable high-frequency robustness and acceptable sensitivity to sensor noise, good stability margins
and almost optimal performance. These rules are also
compared to a set of well established tuning rules.
Finally it is shown that also for plants with time delay, derivative action can oer signicantly improved
performance as long as the time delay is not extremely
large.

Acknowledgments
Valuable discussions and comments on the manu strom and professor
script by professor Karl-Johan A
Sigurd Skogestad are very much appreciated.

References
[1] B. Kristiansson, B. Lennartson, Optimal PID controllers
including rollo and Smith predictor structure, in: 14th World
Congress of IFAC, vol. F, Beijing, China, July 1999, pp. 297302.
[2] B. Kristiansson, PID controllers, design and evaluation, PhD
thesis, Control and Automation Laboratory, Department of
Signals and Systems, Chalmers University of Technology, Goteborg, Sweden, August 2003.
[3] B. Kristiansson, B. Lennartson, Robust and optimal tuning of PI
and PID controllers, IEE Proc. Control Theory Appl. 149 (1)
(2002) 1725.
[4] F.G. Shinskey, Process Control Systems, Application, Design and
Tuning, McGraw-Hill, 1996.
strom, T. Hagglund, The future of PID control, in: PID00,
[5] K.J. A
IFAC Workshop of Digital Control, Past, Present and Future of
PID Control, Terrassa, Spain, April 2000, pp. 1930.
[6] B. Kristiansson, B. Lennartson, Simple and robust tuning of PI
and PID controllers, IEEE Control Syst. Mag., submitted for
publication.
[7] B. Kristiansson, B. Lennartson, Convenient, almost optimal and
robust tuning of PI and PID controllers, in: Proc. of the 15th
World Congress of IFAC, Barcelona, Spain, July 2002.
[8] S. Skogestad, Simple analytic rules for model reduction and PID
controller tuning, J. Process Control 13 (2003) 291309.
strom, T. Hagglund, PID Controllers: Theory, Design and
[9] K.J. A
Tuning, Instrument Society of America, 1995.
[10] W.K. Ho, K.W. Lim, C.C. Hang, L.Y. Ni, Getting more phase
margin and performance out of PID controllers, Automatica 35
(1999) 15791585.
[11] Q.G. Wang, C.C. Hang, Y. Yang, J.B. He, Quantitative robust
stability analysis and PID controller design, IEE Proc. Control
Theory Appl. 149 (1) (2002) 37.
[12] W.K. Ho, W. Xu, PID tuning for unstable processes based on
gain- and phase-margin specications, IEE Proc. Control Theory
Appl. 145 (5) (1998) 392396.
[13] J. Crowe, M.A. Johnson, Towards autonomous PI control
satisfying classical robustness specications, IEE Proc. Control
Theory Appl. 149 (1) (2002) 2631.
[14] R. Gorez, P. Klan, Nonmodel-based explicit design relations for
PID controllers, in: PID00, IFAC Workshop of Digital Control,
Past, Present and Future of PID Control, Terrassa, Spain, April
2000, pp. 141148.

102

B. Kristiansson, B. Lennartson / Journal of Process Control 16 (2006) 91102

[15] F. Morilla, S. Dormido, Methologies for the tuning of PID


controllers in the frequency domain, in: PID00, IFAC Workshop
of Digital Control, Past, Present and Future of PID Control,
Terrassa, Spain, April 2000, pp. 155160.
[16] J. Langer, I.D. Landau, Combined pole placement/sensitivity
function shaping method using convex optimization criteria,
Automatica 35 (6) (1999) 11111120.
[17] P. van Overschee, B.D. Moor, Optimal PID control of a chemical
batch reactor, in: Proc. of the European Control Conference,
Karlsruhe, Germany, 1999, p. CA-1.
strom, T. Hagglund, Design of PID
[18] H. Panagopoulos, K.J. A
controllers based on constrained optimisation, IEE Proc. Control
Theory Appl. 149 (1) (2002) 3240.
[19] T.S. Schei, Automatic tuning of PID controllers based on transfer
function estimation, Automatica 30 (12) (1994) 19831989.
[20] H. Panagopoulos, PID control, design, extension, application, Doctoral thesis, Department of Automatic Control,
Lund Institute of Technology, Lund, Sweden, December
2000.
[21] B. Kristiansson, Evaluation and tuning of PID controllers,
Licentiate thesis, Department of Signals and Systems, Chalmers
University of Technology, Goteborg, Sweden, May 2000.
[22] X.L. Peng, B. Lennartson, C.M. Fransson, On the choice of
sampling period and robust pole placement, in: Proc. of the 15th
World Congress of IFAC, Barcelona, Spain, July 2002.
[23] R. Vilanova, I. Serra, C. Pedret, A general approach to PID
optimal design: achievable designs and generation of stabilizing
controllers, in: PID00, IFAC Workshop of Digital Control, Past,
Present and Future of PID Control, Terrassa, Spain, April 2000,
pp. 444449.
[24] C.A. Neto, M. Embiruco, Tuning of PID controllers: an optimization-based method, in: PID00, IFAC Workshop of Digital
Control, Past, Present and Future of PID Control, Terrassa,
Spain, April 2000, pp. 415420.
[25] G.P. Liu, S. Daley, Optimal-tuning PID controller design in the
frequency domain with application to a rotary hydraulic system,
Control Eng. Practice 7 (1999) 821830.
[26] J.G. Ziegler, N.B. Nichols, Optimum settings for automatic
controllers, Trans. ASME 64 (11) (1942) 759768.
strom, T. Hagglund, A frequency domain method for
[27] K.J. A
automatic tuning of simple feedback loops, in: Proc. of 23rd IEEE
Conf on Decision and Control, Las Vegas NV, USA, December
1984, pp. 299304.
[28] R.J. Mantz, E.J. Tacconi, Complementary rules to Ziegler and
Nichols rules for regulating and tracking controller, Int. J.
Control 49 (1989) 14651471.

strom, T. Hagglund, New tuning methods for PID


[29] K.J. A
controllers, in: Proc. of 3rd European Control Conference, Rome,
Italy, September 1995, pp. 24562462.
[30] P. Persson, Towards autonomous PID control, PhD thesis,
Department of Automatic Control, Lund Institute of Technology,
Lund, Sweden, 1992.
[31] R.R. Pecharroman, F.L. Pagola, Control design for PID controllers auto-tuning based on improved identication, in: PID00,
IFAC Workshop of Digital Control, Past, Present and Future of
PID Control, Terrassa, Spain, April 2000, pp. 8994.
[32] B. Kristiansson, B. Lennartson, Near optimal tuning rules for PI
and PID controllers, in: PID00, IFAC Workshop of Digital
Control, Past, Present and Future of PID Control, Terrassa,
Spain, April 2000, pp. 369374.
strom, W.K. Ho, Renements of the Ziegler
[33] C.C. Hang, K.J. A
Nichols tuning formula, IEE Proc. D, Control Theory Appl. 138
(2) (1991) 111118.
strom, T. Hagglund, Benchmark systems for PID control,
[34] K.J. A
in: PID00, IFAC Workshop of Digital Control, Past, Present and
Future of PID Control, Terrassa, Spain, April 2000, pp. 181182.
strom, H. Panagopoulos, T. Hagglund, Design of PI
[35] K.J. A
controllers based on non-convex optimization, Automatica 34 (5)
(1998) 585601.
[36] G.F. Franklin, J.D. Powell, A. Emami-Naeini, Feedback Control
of Dynamic Systems, third ed., Addison-Wesley, 1994.
strom, Limitations on control system performance, Eur. J.
[37] K.J. A
Control 6 (2000) 220.
strom, Dominant pole designa unied view of
[38] P. Persson, K.J. A
PID controller tuning, in: Preprints 4th IFAC Symposium on
adaptive Systems in Control and Signal Processing, Grenoble,
France, 1992, pp. 127132.
[39] D.E. Rivera, M. Morari, S. Skogestad, Internal model control. 4.
PID controller design, Ind. Eng. Chem. Process Des. Dev. 25
(1986) 252265.
[40] M. Morari, E. Zariou, Robust Process Control, Prentice-Hall,
Englewood Clis, New Jersey, 1989.
[41] A.J. Isaksson, S.F. Graebe, Analytical PID parameter expressions
for higher order systems, Automatica 35 (6) (1999) 11211130.
[42] F.G. Shinskey, How good are our controllers in absolute
performance and robustness? Meas. Control 23 (1990) 114121.
[43] A. Ingimundarson, T. Hagglund, Performance comparison
between PID and dead-time compensating controllers, J. Process
Control 12 (8) (2002) 887895.
strom, T. Hagglund, Revisiting the ZieglerNichols step
[44] K.J. A
response method for PID control, J. Process Control 14 (2004)
635650.

You might also like