J. Cent. South Univ.

(2014) 21: 3777−3786

DOI: 10.1007/s11771-014-2362-0

Robust PID controller design for time delay processes with peak of
maximum sensitivity criteria

Mohammad Shamsuzzoha
Department of Chemical Engineering, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia
© Central South University Press and Springer-Verlag Berlin Heidelberg 2014

Abstract: The motivation of this work is to obtain single PI/PID tuning formula for different types of processes with enhanced
disturbance rejection performance. The proposed tuning formula consistently gives better performance in comparison to several
well-known methods at the same degree of robustness for stable, integrating and unstable processes. For the selection of the
closed-loop time constant (τc), a guideline is provided over a broad range of time-delay/time-constant ratios on the basis of the peak
of maximum sensitivity (Ms). An analysis has been performed for the uncertainty margin with the different process parameters for the
robust controller design. It gives the guideline of the Ms-value settings for the PI controller designs based on the process parameters
uncertainty. Furthermore, a relationship has been developed between Ms-value and uncertainty margin with the different process
parameters (k, τ and θ). Simulation study has been conducted for the broad class of processes and the controllers are tuned to have the
same degree of robustness by measuring the maximum sensitivity, Ms, in order to obtain a reasonable comparison.

Key words: PI/PID controller tuning; internal model control (IMC) method; unstable delay process; integrating delay process;
disturbance rejection

procedure. Due to this reason, a large number of PID

1 Introduction
The proportional integral derivative (PID) controller
is the most widely used control algorithm for regulatory
control layer in process industries. It is because of their
simplicity, ease of implementation, robustness and wide
ranges of applicability with near-optimal performance.
The stable and integrating processes are very common in
process industries in flow, level and temperature loop.
The open-loop unstable processes are also encountered in
chemical processing units and known to be difficult to
control, especially when there exists a time delay, such as
in the case of continuous stirred tank reactors,
polymerization reactors and bioreactors which are
sometimes open-loop unstable by design. A recent survey
of KANO and OGAWA [1] shows that the ratio of
applications of different types of controller, e.g., PI
control, conventional advanced control and model
predictive control is about 100:10:1. There is no perfect
alternative of the PID controller at least at the bottom
layer in the process industries. This was the clear
conclusion at the end of the IFAC Conference on
Advances in PID Control, held in Brescia (Italy) during
28−30 March, 2012. Although the PID controller has
only three adjustable parameters, it is not easy to get
suitable controller settings without a systematic
controllers are poorly tuned in industries.
There are variety of controller tuning approaches
reported in the literature and among that two types are
widely used, one may use open-loop or closed-loop plant
test. Most tuning approaches are based on open-loop
plant information; typically the plant’s gain (k), time
constant (τ) and time delay (θ).
The effectiveness of the internal model control
(IMC) design principle has made it attractive in the
process industries, where many attempts have been made
to exploit the IMC principle to design PI/PID controllers
for both stable and unstable processes. The IMC-PID
tuning rules have the advantage of using only a single
tuning parameter to achieve a clear trade-off between the
closed-loop performance and robustness. The PI/PID
tuning methods proposed by RIVERA et al [2], LEE et al
[3], SKOGESTAD [4] and SHAMSUZZOHA and
LEE [5], are typical examples of the IMC-PID tuning
method. The direct synthesis (DS) method [6] and the
direct synthesis for the disturbance (DS-d) method
proposed by CHEN and SEBORG [7] can also be
categorized into the same class as the IMC-PID methods,
in that they obtain the PI/PID controller parameters by
computing the ideal feedback controller which gives a
predefined desired closed-loop response. Although the
ideal controller is often more complicated than the PI/PID

Received date: 2013−12−30; Accepted date: 2014−05−29

Corresponding author: Mohammad Shamsuzzoha, Assistant Professor, PhD; Tel: +966−138607360; E-mail: mshams@kfupm.edu.sa
3778
controller for time delayed processes, the controller form
can be reduced to that of either a PI/PID controller or a
PID controller cascaded with a low order filter by
performing appropriate approximations of the dead time
in the process model.
It is essential to emphasize that the PI/PID
controller designed according to the IMC principle
provides excellent set-point tracking, but has a sluggish
disturbance response, especially for processes with a
small time-delay/time-constant ratio [2−5, 8]. In process
industries, disturbance rejection is much more important
than set-point tracking for many process control
applications, a controller design that emphasizes the
former rather than the latter is an important design
problem and it has been the focus of the research.
It is well known that the IMC structure is very
powerful for controlling stable processes with time delay
and cannot be directly used for unstable processes
because of the internal instability [8]. For this reason,
some modified IMC methods of two-degree-of-freedom
(2DOF) control were developed for controlling unstable
processes with time delay, such as those proposed by
LEE et al [9], YANG et al [10], WANG and CAI [11],
TAN et al [12], LIU et al [13] and JUNG et al [14]. In
addition, 2DOF control methods based on the
Smith-Predictor (SP) were proposed by MAJHI and
ATHERTON [15], KWAK et al [16], and ZHANG et al
[17] to achieve a smooth nominal setpoint response
without overshoot for first order unstable processes with
time delay.
The delay integrating process is very important in
process industries. It has a clear advantage in the
identification test, because the model contains only two
parameters and is simple to use for identification. Some
of the well accepted PI/PID controllers tuning methods
for delay integrating processes are those proposed by
SKOGESTAD [4], CHEN and SEBORG [7] and TYRUS
and LUBYEN [18], SHAMSUZZOHA and LEE [19].
ALCANTARA et al [20] have addressed the
model-based tuning of the PI/PID controller based on the
robustness/performance and servo/regulator trade-offs.
K-SIMC method, a modification of SIMC rule has been
proposed by LEE et al [21] for stable process and
subsequently for unstable process [22]. TORRICO et al
[23] proposed a new and simple design for the filtered
Smith predictor (FSP), which belongs to a class of
dead-time compensators (DTCs) and allows the handling
of stable, unstable, and integrating processes. ALFARO
and VILANOVA [24] have proposed unified simple
optimal and robust tuning (uSORT1) method. It is 1DOF
PI/PID controller tuning method for the FOPDT and
SOPDT process. The uSORT1 method allows to adjust
the control system robustness varying only the controller
gain. JIN and LIU [25] proposed IMC-PI tuning rules
J. Cent. South Univ. (2014) 21: 3777−3786
based on model matching approach and closed-loop
shaping. Subsequently, JIN and LIU [26] suggested
analytical tuning rules of 2-Degree of freedom PID
controller for integrating processes, which was based on
an enhanced IMC principle for IPTD, IFOPTD and
DIPTD models with robustness/performance criteria.
JENG and LIN [27] proposed robust PID tuning for
stable/integrating processes with inverse response and
time delay. Their design method is based on a Smith-type
compensator for nonminimum phase dynamics. Based on
the optimization problem, ARRIETA and VILANOVA
[28] addressed the combined servo/regulation
performance and robustness problem for PID controller
tuning. A unified approach was proposed by
NORMEY-RICO and GUZMÁN [29], which is based on
a PID approximation of the filtered Smith predictor, for
tuning PID controllers for stable, integrating, and
unstable dead-time processes.
The main alternative of the above mentioned
open-loop approach is to use closed-loop experiments.
One approach is the classical method of Ziegler-Nichols
[30] which requires very little information about the
process to obtain controller setting. Recently, several
authors [31−33] have proposed modified tuning methods
based on closed-loop experiments and resulting
controller gives better performance. The recent published
online controller tuning method in closed-loop mode by
SHAMSUZZOHA [33] overcomes the shortcoming of
the well-known Ziegler-Nichols continuous cycling
method and gives consistently better performance and
robustness for a broad class of processes.
It should be emphasized that the design principle of
the most of the aforementioned tuning methods is
complicated and that the modified IMC structure for
unstable process is difficult to implement in a real
process plant in the presence of model uncertainty.
Therefore, in this work, a simple analytical method
is proposed for the design of the PI/PID controller.
Overcoming the drawback of other tuning rules for
different types of processes, only single tuning rule is
capable to handle different types of processes with
performance improvement. A τc guideline was
recommended for a wide range of time-delay/time-
constant ratios (θ/τ). A guideline of the Ms-value settings
has been also proposed based on the process parameters
uncertainty (k, τ and θ) margin. Simulation study was
performed to show the validity of the proposed method at
same Ms-value.
2 Theory of IMC-PID controller design
Figures 1(a) and (b) show the block diagrams of the
IMC control and equivalent classical feedback control
structures, respectively, where Gp is the process, G p is
J. Cent. South Univ. (2014) 21: 3777−3786 3779
the process model, q is the IMC controller, fr is the Then, the IMC controller comes to be
set-point filter, and Gc is the equivalent feedback m
controller. (  i s i  1)
q  pm1 i 1
(5)
( c s  1) r
Thus, the resulting set-point and disturbance
responses are obtained as
m

y
 i si  1
 Gp qf r  pA i 1 fr (6)
r ( c s  1) r

 m 
y
   i si  1 
 (1  Gp q )Gp  1  pA i 1 G (7)
d  ( c s  1) r 
p
 
 
m
Fig. 1 Block diagram of IMC and classical feedback control
systems: (a) IMC structure; (b) Feedback control structure
The numerator expression  i si  1 in Eq. (6)
i 1
causes an excessive overshoot in the servo response,
For the nominal case Gp  G p , the set-point and which can be eliminated by introducing the set-point
disturbance responses in the IMC control structure can filter fr to compensate for the overshoot in the servo
be simplified as response.
y  Gp qf r r  (1  G p q)Gp d (1) From the above design procedure, a stable,
closed-loop response can be achieved by using the IMC
According to the IMC parameterization [2], the controller. The ideal feedback controller that is
process model G p is factored into two parts: equivalent to the IMC controller can be expressed in
G p  pm pA (2) terms of the internal model G p and the IMC controller
q as:
where pm is the portion of the model inverted by the
controller, pA is the portion of the model not inverted by q
Gc  (8)
the controller and pA(0)=1. The noninvertible part 1  G p q
usually includes the dead time and/or right half plane
Substituting Eqs. (2) and (5) into Eq. (8) gives the
zeros and is chosen to be all-pass.
ideal feedback controller:
To get a superior response for unstable processes or
m
stable processes with poles near zero, the IMC controller
(  i s i  1)
q should satisfy following conditions. If the process Gp
pm1 i 1
has unstable poles or poles near zero at z1, z2,…, zm, then ( c s  1) r
Gc  (9)
q should have zeroes at z1, z2,…, zm, and (1−Gpq) should m
pA (  i s  1) i
also have zeroes at z1, z2,…, zm.
1 i 1
Since the IMC controller q is designed as ( c s  1) r
q  pm1 f , the first condition is satisfied automatically.
The second condition can be fulfilled by designing the The resulting controller in Eq. (9) is physically
IMC filter (f) as realizable but it does not have the standard PI/PID form.
The desired form of the controller can be obtained by
im1  i s i  1
f  (3) using the proper approximation of the dead time term for
( c s  1) r example Taylor series expansion. In this work, both
where τc is an adjustable parameter which controls the simplicity and approximation error due to dead time term
tradeoff between the performance and robustness; r is has been considered carefully during the PI/PID
selected to be large enough to make the IMC controller controller design.
(semi-)proper; αi is determined by Eq. (4) to cancel the
poles near zero in Gp. 3 PI/PID controller design
pA (im1  i s i  1)
1  Gp q  1 0 (4) This section illustrates the PI/PID controller design
s  z1 , zm ( c s  1) r sz ,, zm method for several representative cases which are
1
3780 J. Cent. South Univ. (2014) 21: 3777−3786
frequently used by the process control engineer in real process becomes the series-form of the controller as
processes.  1  Ds 1 
Gc  K c  1    (17)
  Is  D N s 1 
3.1 First-order plus dead time process
First order plus dead time (FOPDT) process is where τD is the derivative time. In the simulation study,
representative model and commonly used in the chemical filter parameter N is typically around 100 and can be
process industries. utilized to make the series-PID controller with derivative
filter. The implementation of the series-PID structure in
ke s
Gp  (10) Eq. (17) is shown in Fig. 2. In the simulation of the
 s 1 second order process, the robustness margins have been
where k is the process gain, τ is the time constant, and θ computed with τD/N=0. The PID setting of the proposed
is the time delay, the IMC filter structure selected is method is for the series form of the PID controller. If
 s 1 required, it can be easily converted to the ideal (parallel)
f  (11)
( c s  1) 2 form of the PID.

After utilizing the above IMC design principle the

ideal feedback controller is given as
( s  1)( s  1)
Gc  (12)
k ( c s  1) 2  e  s ( s  1) 
 
From Eq. (12), the resulting PI controller can be
obtained using Taylor series expansion, e−θs=1−θs and
Fig. 2 Cascade implementation of PID controller without
then simplified as
differentiation of setpoint

Kc  ; I   (13)
k (2 c     ) Here, we should note the follows. First-order
Furthermore, it is obvious that the remaining part of delayed unstable process (FODUP) does not have the
the denominator in Eq. (12) contains the factor of the form of Eq. (10). It can be easily transformed to the form
process poles (τs+1). It has been ignored because of its of Eq. (10) by adjusting their sign for PI controller
little impact on the control performance, while keeping design, e.g., for the FODUP -k and -τ. Similar concept is
the simple PI control structure. also applicable for the second order unstable process
The value of α is selected so that it cancels out the with time delay. The delayed integrating process (DIP)
pole at s=−1/τ. From Eq. (4), this requires [1−(αs+1)e−θs/ can be modeled by considering the integrator as a stable
(τcs+1)2]s=−1/τ and the value of α is obtained as pole near zero. This is mandatory since it is not
practicable to implement the aforementioned IMC based
   c    
2
design procedure for the DIP, because the term, α,
   1  1  e  (14)
     vanishes at s=0.
As a result, the DIP can be approximated to the
FOPDT as
3.2 Second-order plus dead time process
Consider a stable second-order plus dead time ke  s ke  s  ke  s
Gp    (18)
(SOPDT) process as s s  1/  s 1
where ψ is a constant with a sufficiently large value, e.g.,
ke  s
Gp  (15) ψ=100. Accordingly, the optimum IMC filter structure
( s  1)( 2 s  1) for the DIP is identical to that of the FOPDT and
The recommended controller setting for the SOPDT resulting PI parameters is same as FOPDT.
process is PID. It is mainly recommended for the
“dominant” second-order process. It means that the 3.3 Setpoint filter to enhance servo response
second-order time constant (τ2) is larger than the In the proposed controller design, term (αs+1)
effective time delay θ, i.e., τ2>θ. shows a large overshoot for the step set-point change, it
Based on the IMC design principle, the PID is because the controller is designed based on the
controller setting for the SOPDT process is given as disturbance rejection. Therefore, a set-point filter fr is
suggested to remove excessive overshoot and enhance
 the servo response.
Kc  ; I  ; D  2 (16)
k (2 c     )  s 1
fr  c (19)
The proposed PID controller setting for the SOPDT ( s  1)
J. Cent. South Univ. (2014) 21: 3777−3786
TAN et al [12] also suggested this form of set-point
filter. Due to this type of lead-lag filter, resulting
response will be first order with the time constant of τc
for the set-point change.
4 Simulation study
Although the simulation study is conducted on the
different types of processes, only few of them are
discussed below. The proposed tuning rule provides
acceptable controller settings in all cases with respect to
both performance and robustness. The closed-loop
performance is evaluated by introducing a unit step
change in both the set-point and load disturbance, i.e.,
(ys=1 and d=1). The brief overview of the performance
and robustness measure is mentioned below.
Output performance (y) is quantified by computing

the integrated absolute error, EIA = 
0
Manipulated variable usage is quantified by calculating
the total variation (VT) of the input (u), which is the sum
of all its moves up and down. If input signal is
discretized as a sequence [u1, u2, u3…, ui…], then

VT =  ui +1  ui . Note also that VT is the integral of the
i=1
absolute value of the derivative of the input,
 du
VT =  dt , so V T is a good measure of the
0 dt
smoothness. To evaluate the robustness, maximum
closed-loop sensitivity is computed in the present work
which is defined as Ms=maxω|1/[1+gc(jω)]. Since Ms is
3781
the inverse of the shortest distance from the Nyquist
curve of the loop transfer function to the critical point
(−1, 0), a small Ms-value indicates that the control
system has a large stability margin. It is better to have
EIA, VT and Ms all to be small, but for a well-tuned
controller there is a trade-off, which means that a
reduction in EIA implies an increase in VT and Ms, (and
vice versa).
To achieve the fair comparisons in the simulation
study, all controllers have been tuned by adjusting τc for
the same degree of robustness by fixing Ms. The results
of five different types of processes with performance and
robustness matrix are listed in Table 1.
Figures 3−7 show comparison of the proposed
method with other methods like SIMC [4], DCLR [3]
and TL [18]. In case of stable and integrating process
proposed method gives faster disturbance rejection and
y  y s dt . has clear advantage over the DCLR and SIMC methods.
The proposed method also works well in first and second
order unstable processes with dead time. The results of
examples C4 clearly show that the proposed method
gives both smaller overshoot and faster disturbance
rejection while maintaining setpoint performance for
unstable process. From above analysis, it seems that the
proposed method constantly gives better closed-loop
response for several types of processes at same Ms-value
compared with other methods.
For the PI-controller design, delay integrating
process (DIP) should be approximated to the FOPDT
process. In the present simulation case, C2 has been
modeled as 20e−7.4s/(100s+1).

Table 1 PI/PID controller setting for proposed and other methods with performance matrix
Setpoint Load disturbance
Case Process Method τc Ms Kc τI VT
EIA: (y) EIA (y) VT
Proposed 2.46 1.60 4.57 4.85 3.1 5.9 1.06 1.37
e s
C1 SIMC 1.0 1.60 5.0 8.0 2.5 5.6 1.60 1.16
10s  1
LEE et al [3] 1.0 1.60 5.12 10.25 2.17 5.58 2.0 1.10
Proposed 19.37 1.70 0.304 39.63 30.21 0.51 131.9 1.74
0.2e7.4s
C2 SIMC 7.4 1.70 0.338 59.2 28.8 0.49 174.5 1.55
s
TL — 1.67 0.33 64.7 29.13 0.46 195 1.50

e 0.5 s Proposed 1.36 6.0 1.646 8.25 6.72 12.0 5.01 7.30
C3
s 1 LEE et al [9] 1.4 6.0 1.668 8.67 6.77 12.04 5.20 7.24
0.5 s Proposed 2.9 2.2 3.22 9.50 6.37 7.59 2.95 2.59
e
C4
(5s  1)(2 s  1)(0.5s  1) YANG et al [10] 1.5 2.2 2.564 10.98 8.57 7.24 4.28 2.95

10e  s Proposed 2.57 1.65 0.942 5.55 20.39 1.48 5.89 2.44
C5
(20 s  1)(20 s  1) SIMC 1.0 1.65 1.0 8.0 20.6 1.44 8.0 2.44
Note: For case C4, τD=2.25 and 1.82 is used for proposed and YANG et al [10] methods, respectively; For case C5, τD=20 is used for both proposed and SIMC
methods, response is without setpoint filter for both the methods. A set-point filter fr is suggested to enhance servo response, for PI controller
f r  ( c s  1) /( s  1); for PID f r  ( c s  1) /( 2 s 2   s  1).
3782
Fig. 3 Responses of PI-control of first-order process Gp=
e−s/(10s+1) (C1) for both setpoint (a) and load disturbance (b)
of magnitude 1 at t=0
Figure 6 shows the manipulated variable (MV)
response for C1 as the representative case. The response
of the MV of the proposed method is comparable with
the SIMC [4] and LEE et al [3].
4.1 Proposed method for dominant second order
processes
A published SOPDT model [20] which has τ1≈τ2
was considered for the performance comparison as
10e  s
C5 (SOPDT): Gp 
(20 s  1)(20 s  1)
The proposed method has been compared with the
SIMC method for C5. The SIMC parameters of the
PID-series controller settings has been obtained for
τc=θ=1, which gives Ms=1.65. In order to ensure a fair
comparison, the proposed PID setting is tuned to have
same Ms=1.65 by adjusting their closed loop time
constant, τc=2.57. To compare the response, a unit step
change is introduced in both the set-point and load
disturbance. Figure 7 compares the set-point and load
disturbance responses obtained using both the compared
controllers. The closed loop response for setpoint change
J. Cent. South Univ. (2014) 21: 3777−3786
Fig. 4 Responses of PI-control of integrating process
Gp=0.2e−7.4s/s (C2) for both setpoint (a) and load disturbance (b)
of magnitude 1 at t=0
for both the proposed and SIMC method are almost
identical, while there is significant advantage of the
proposed method in disturbance rejection where it shows
smaller overshoot and fast settling time.
4.2 Effect of setpoint filter on servo response
The proposed method is based on the disturbance
rejection so the large overshoot for the step setpoint
change can occur particularly for the unstable and
integrating processes. Therefore, lead-lag set-point filter
is recommended to remove the overshoot in setpoint
response. To show the performance improvement, a first-
(20)
order unstable process with time delay (Case C3) has
been considered. The resulting set-point filter of the
proposed study for C3 should be fr=(1.36s+1)/(8.25s+1).
Figure 8 shows the closed-loop response of the
proposed method for both with and without set-point
filter where EIA is reduced from 6.72 to 1.90. As
expected the output response with set-point filter is fast
without any overshoot.
5 Discussion
In the proposed tuning rule, the closed-loop time
constant τc controls the tradeoff between robustness and
J. Cent. South Univ. (2014) 21: 3777−3786
Fig. 5 Responses of PID-control for high order unstable
process Gp=e−0.5s/(5s−1)(2s+1)(0.5s+1) (C4) for both setpoint
(a) and load disturbance (b) of magnitude 1 at t=0
performance of the control system. As τc decreases, the
closed-loop response becomes faster and can be unstable.
On the other hand, as τc increases, the closed-loop
response becomes sluggish and more stable. A good
tradeoff is obtained by choosing τc to give Ms-value in
the range of 1.2−2.0 for stable process. An analysis of
the τc selection has been conducted and plot of Ms verses
θ/τ for different values of τc= γθ, where γ=1.0, 2.0 and
3.0 is shown in Fig. 9. The figure clearly shows that τc=θ
is not the proper choice because for the lag dominant
process, it gives tight controller setting. For τc=3θ, it
gives smooth and robust setting because Ms lies between
1.61 to 1.25. A good tradeoff between robustness and
performance can be achieved for τc=2θ where it gives
Ms=2.0 for lag dominant process and Ms=1.26 for delay
dominant process.
6 Robustness study
This section presents the analysis of the control
system design for system affected by parametric
uncertainty. It indicates the maximum uncertainty margin
in different process parameters for the fixed Ms-value. It
3783
Fig. 6 MV plots of PI-control for first-order process e−s/(10s+1)
(C1) for both setpoint (a) and load disturbance (b) of magnitude
1 at t=0
is important to obtain the relationship between Ms-value
and parametric uncertainty in the control system design.
It is because these uncertainties play an important role in
control system and cause poor performance or even
instability of closed-loop control systems. A typical first
order delay process (e−θs/(10s+1)) is considered for this
analysis for various dead time to lag time ratios by
changing θ while fixing τ=10). The Kharitonov’s
theorem is used to obtain the uncertainty margin in the
process parameter and further it is verified by the
simulation for different θ/τ ratio. The percentages of the
uncertainty margin in different parameters have been
analyzed for different Ms-value. Figure 10 shows the
variation in dead time margin for different Ms-values.
The figure clearly indicates that for a fixed Ms-value, as
the θ/τ ratio increases, the dead time margin also
increases. The variation in the process gain uncertainty
for the different Ms-values is shown in Fig. 11. The trend
in Fig. 11 is reverse and it shows that for the fixed
Ms-value, as θ/τ ratio increases, the percentage in the
gain margin decreases. The uncertainty in the process
time constant τ is shown in Fig. 12. The uncertainty
margin in the τ is lower than the original value whereas
3784
Fig. 7 Responses of PID-control of second-order process
10e  s
Gp= (C5) for both setpoint (a) and load
(20 s  1)(20 s  1)
disturbance (b) of magnitude 1 at t=0
Fig. 8 Effect of setpoint filter to remove overshoot from
setpoint response: Setpoint responses of PI-control of first-
order unstable process with time delay Gp=e−0.5s/(s−1) (C3) for
both setpoint and load disturbance of magnitude 1 at t=0
in the k and θ is higher than original values. These
combinations have deteriorating impact on the control
system performance. The maximum tolerance limit for
J. Cent. South Univ. (2014) 21: 3777−3786
Fig. 9 τc guidelines of first order stable process with time delay
based on time delay in process
Fig. 10 Variation of uncertainty margin in time delay (θ) with
θ/τ ratio for different Ms-values
Fig. 11 Variation of uncertainty margin in process gain (k) with
θ/τ ratio for different Ms-values
wide range of θ/τ ratio of the uncertainty in different
process parameters (k, θ and τ) is given in the Table 2.
Based on this information, one can select the suitable
Ms-value for safe PI controller design of the uncertain
process. The other PI tuning methods have almost similar
combinations of the uncertainly margin for fixed Ms-value.
J. Cent. South Univ. (2014) 21: 3777−3786 3785
time constant is τc=2θ for a wide range of θ/τ ratio.
7) The guideline of the Ms-value is given for the PI
controller design of the uncertain process. The proposed
investigation of the Ms-value verses uncertainty margin
in the process parameter can be very useful for the robust
controller design.
8) Several important representative processes are
considered in the simulation study in order to
demonstrate the advantage of the proposed method. In
particular, the proposed controller shows excellent
performance when the lag time dominates.

Acknowledgement
The author would like to acknowledge the support
Fig. 12 Variation of uncertainty margin in process time constant
provided by King Abdulaziz City for Science and
(τ) with θ/τ ratio for different Ms-values Technology (KACST) through the “KACST Annual
Program” at King Fahd University of Petroleum &
Table 2 Maximum uncertainty margin in θ, k and τ for different
Minerals (KFUPM) for funding this work through
value of Ms project number AT-32-41.
Uncertainty Uncertainty Uncertainty
Ms
margin in θ/% margin in k/% margin in τ/%
References
1.4 278 322 79.4
[1] KANO M, OGAWA M. The state of art in chemical process control
1.6 183 211 72.0 in Japan: Good practice and questionnaire survey [J]. Journal of
1.8 136 157 66.1 Process Control, 2010, 20: 969−982.
[2] RIVERA D, MORARI M, SKOGESTAD S. Internal model control. 4.
2.0 108 126 61.2 PID controller design [J]. Ind Eng Chem Process Des Dev, 1986, 25:
252−265.
7 Conclusions [3] LEE Y, PARK S, LEE M, BROSILOW C. PID controller tuning for
desired closed–loop responses for SI/SO systems [J]. AIChE, 1998,
44: 106−115.
1) A simple analytical design method for the PI/PID [4] SKOGESTAD S. Simple analytic rules for model reduction and PID
controller is proposed, which is based on the IMC controller tuning [J]. Journal of Process Control, 2003, 13: 291−309.
principle in order to improve the disturbance rejection [5] SHAMSUZZOHA M, LEE M. IMC–PID controller design for
performance. improved disturbance rejection of time–delayed processes [J]. Ind
Eng Chem Res, 2007, 46: 2077−2091.
2) Another important feature of the proposed
[6] SMITH C L, CORRIPIO A B, MARTIN J. Controller tuning from
methodology is that it deals stable, integrating and simple process models [J]. Instrum Technol, 1975, 22(12): 39
unstable process in a unified way. [7] CHEN D, SEBORG D. PI/PID controller design based on direct
3) In conclusion, the final tuning formula for the synthesis and disturbance rejection [J]. Industrial and Engineering
proposed PI/PID tuning rule is summarized as Chemistry Research, 2002, 41: 4807−4822.
[8] MORARI M, ZAFIRIOU E. Robust process control [M]. New Jersey:

Kc  ;  I   ;  D   2 ; and Prentice-Hall Englewood Cliffs, 1989: 85−111.
k (2 c     ) [9] LEE Y, LEE J. PARK S. PID controller tuning for integrating and
unstable processes with time delay [J]. Chemical Engineering
   c    
2
Science, 2000, 55: 3481−3493.
   1  1  e 
     [10] YANG X, WANG Q, HANG C, LIN C. IMC-based control system
design for unstable processes [J]. Industrial and Engineering
4) For the first order and integrating process with Chemistry Research, 2002, 41: 4288−4294.
[11] WANG Y, CAI W. Advanced proportional-integral-derivative tuning
time delay, the resulting tuning rule is PI where τD=0.
for integrating and unstable processes with gain and phase margin
5) The design method is based on the disturbance specifications [J]. Industrial and Engineering Chemistry Research,
rejection and a set-point filter is recommended to 2002, 41: 2910−2914.
eliminate the overshoot in set-point response mainly in [12] TAN W, MARQUEZ H, CHEN T. IMC design for unstable processes
unstable and integrating process. with time delays [J]. Journal of Process Control, 2003, 13: 203−213.
[13] LIU T, ZHANG W, GU D. Analytical design of
6) For a given process model, parameter τc is a
two-degree-of-freedom control scheme for open-loop unstable
single adjustable parameters in the tuning rule and it is
process with time delay [J]. Journal of Process Control, 2005, 15:
used to obtain a compromise between performance and 559−572.
robustness. The recommended selection of closed loop [14] JUNG C S, SONG H K, HYUN C J. A direct synthesis tuning
3786
method of unstable first-order-plus-time-delay processes [J]. Journal
of Process Control, 1999, 9: 265−269.
[15] MAJHI S, Atherton D. Obtaining controller parameters for a new
smith predictor using autotuning [J]. Automatica, 2000, 36:
1651−1658.
[16] KWAK H, SUNG S, LEE I B, PARK J. Modified smith predictor
with a new structure for unstable processes [J]. Industrial and
Engineering Chemistry Research, 1999, 38: 405−411.
[17] ZHANG W, GU D, WANG W, XU X. Quantitative performance
design of a modified smith predictor for unstable processes with time
delay [J]. Industrial and Engineering Chemistry Research , 2004, 43:
56−62.
[18] TYREUS B, LUYBEN W. Tuning PI controllers for integrator/dead
time processes [J]. Ind Eng Chem Res, 1992: 2625−2628.
[19] SHAMSUZZOHA M, LEE M. Analytical design of enhanced
PID·filter controller for integrating and first order unstable processes
with time delay [J]. Chemical Engineering Science, 2008, 63:
2717−2731.
[20] ALCANTARA S, VILANOVA R, PEDRET C. PID control in terms
of robustness/performance and servo/regulator trade-offs: A unifying
approach to balanced autotuning [J]. Journal of Process Control,
2013, 23: 527−542.
[21] LEE J, CHO W, EDGAR T F. Simple analytic PID controller tuning
rules revisited [J]. Industrial & Engineering Chemistry Research,
2014, 53: 5038−5047.
[22] CHO W, LEE L, EDGAR T. Simple analytic
proportional-integral-derivative (PID) controller tuning rules for
unstable processes [J]. Industrial & Engineering Chemistry Research,
2014, 53: 5048−5054.
[23] TORRICO B C, CAVALCANTE M U, BRAGA A P,
NORMEY-RICO J E, ALBUQUERQUE A A. Simple tuning rules
for dead-time compensation of stable, integrative, and unstable
first-order dead-time processes [J]. Industrial & Engineering
Chemistry Research, 2013, 52: 11646−11654.
J. Cent. South Univ. (2014) 21: 3777−3786
[24] ALFARO V M, VILANOVA R. Optimal robust tuning for 1DoF
PI/PID control unifying FOPDT/SOPDT models [C]// in IFAC
Conference on Advances in PID Control PID’12, Brescia (Italy):
IFAC, 2012.
[25] JIN Q B, LIU Q. IMC-PID design based on model matching
approach and closed-loop shaping [J]. ISA Transactions, 2014, 53:
462−473.
[26] JIN Q B, LIU Q. Analytical IMC-PID design in terms of
performance/robustnesstradeoff for integrating processes: From
2-Dof to 1-Dof [J]. Journal of Process Control , 2014, 24: 22−32.
[27] JENG J C, LIN S W. Robust Proportional-integral-derivative
controller design for stable/integrating processes with inverse
response and time delay [J]. Ind Eng Chem Res, 2012, 51:
2652−2665.
[28] ARRIETA O, VILANOVA R. Simple servo/regulation
proportional-integral-derivative (PID) tuning rules for arbitrary
Ms-based robustness achievement [J]. Ind Eng Chem Res, 2012, 51:
2666−2674.
[29] NORMEY-RICO AND J E, GUZMÁN J L. Unified PID tuning
approach for stable, integrative, and unstable dead-time processes [J].
Ind Eng Chem Res, 2013, 52: 16811−16819.
[30] ZIEGLER J G. NICHOLS N B. Optimum settings for automatic
controllers [J]. Trans ASME, 1942, 64: 759−768.
[31] SHAMSUZZOHA M, SKOGESTAD S. The setpoint overshoot
method: A simple and fast closed-loop approach for PID tuning [J].
Journal of Process Control, 2010, 20: 1220−1234.
[32] HU W, XIAO G. Analytical proportional-integral (PI) controller
tuning using closed-loop setpoint response [J]. Ind Eng Chem Res,
2011: 2461-2466.
[33] SHAMSUZZOHA M. Closed-loop PI/PID controller tuning for
stable and integrating process with time delay [J]. Ind Eng Chem Res,
2013, 52: 12973−12992.
(Edited by DENG Lü-xiang)