PID past, present and Future

Prof. Cesar de Prada

Dpt. Systems Engineering and Automatic Control
University of Valladolid, Spain
prada@autom.uva.es
http://www.isa.cie.uva.es/~prada/

CERN, 6 December 2013

1
 Outline

Introduction
PID controller
Types of PID controllers
PID tuning
Automatic tuning
Controller supervision
Aim: Provide an overview of the most relevant topics related
to PID controllers
2
 A control loop
It replaces the human
operator
w FC
u
y
a q

Set point tracking

Flowmeter Valve Disturbance
rejection
Controller
Actuator
w e u y
+
- Computes u Process
Transmitter

3
 Main concepts

Set point or desired value w Automatic / Manual

Controlled variable y Local/Slave
Manipulated variable u, MV, OP ..
Algorithm for computing u
Error
4
 The PID algorithm

e( t ) = w( t ) y( t )
1 de
u( t ) = K p e( t ) + e( )d + Td
Ti dt

Signal based controller, no explicit process

knowledge is incorporated
3 tuning parameters Kp, Ti, Td
Many different implementations

5
 A bit of history
1911 First application of a PID controller by Elmer Sperry.
1920 First patent of a PI controller
1933 - Taylor Double-response plus Fulscope (Model 56R
Fulscope) with adjustable P and I componenets
1925-1935: Widespread use of the PID in industry thanks to the action of
instrumentation companies such as Foxboro and Taylor. 75.000 automatic
controllers sold in the USA

1939 First fully adjustable

commercial controller:
Fulscope 100
from Taylor Instruments

Pneumatic technology
6
 Normalised I/O signals

Actuator
w u y
Controller Process
4-20 mA
4-20 mA Transmitter

4-20 mA from PLC

the transmitter
SP 45
PV 45.5 4-20 mA to the
actuator

M
V
Loop
38 controller
7
The PID
Implementation
algorithm is
implemented as
software in the
DCS controller Control room
modules
Control
modules
Input
4 20 mA /output
modules

Control wardrobe 8
 Digital Control
u(kT)

Computer D/A Process

y(kT)
A/D
T sampling period T

T should be chosen according to the process dynamics, as well as

considering numerical problems in integration and differentiation.
Integration: T 0.1 ...0.3 Ti Differenciation. T 0.2 ...0.6 Td / N
Accuracy in the measurement depends also on the D/A converter
Higher precision in the internal computations than the one of D/A
9
 Digital PID
e( t ) = w ( t ) y ( t )
u( t ) = u( t 1) + g 0 e( t ) + g1 e( t 1) + g 2 e( t 2)

Many formulas for discretization

Microprocesor based controller with many
auxiliary functions
Sampling time T very often fixed in the range
100...200 msg
10
 Architectures

HART I/O DeviceNet/Profibus

AS-i
H1

11
Operation

Typical
PID face

Typical
operator
screen

12
 Configuration

Forms with
configuration
parameters

Type, Units,
constraints SP,
I/O conexions,
man/auto, tuning,
alarms,.

13
 PID actions
e( t ) = w ( t ) y ( t )
1 de
u ( t ) = K p e( t ) + e()d + Td
Ti dt
Kp gain / Proportional term
% span control / % span controlled variable
Proportional band PB=100/ Kp
Ti integral time / Integral term
minutes o sg. (per repetition) (reset time)
repetitions per min = 1/ Ti
Td derivative time / Derivative term
minutes o sg. 14
 Proportional controller P

u ( t ) = K p e( t ) + bias

e u

t t

An error of x % creates an action of

Kp x % on the actuator

bias = manual reset (CV = SP)

15
 Direct / Reverse Acting
e = w-y
Type of valve should be
taken into account
LC w

y
Air open valve

LT LC LT

Direct acting controller Kp < 0 Reverse acting controller Kp > 0

u(t)=Kp(w-y) if y increases, then u decreases if Kp is

positive
16
 Proportional action

1500 w e u
+
Kp Ampl.
rpm -

30 %
Ing. M

1500
rpm

There is only an equilibrium

u(t)=Kp e(t) + 30 point with zero error 17
 Integral action

Kp
Ti ed
w e u
1500 +
Kp Ampl.
-
rpm

Ing. M

1500
rpm
18
 Integral action (automatic reset)

w y w
y

t t

u u
Kp
t Ti ed t

A P controller does not get The integral term changes

steady zero error with self- continuously the control
regulated processes signal until the error is zero 19
 Integral action

t
Kp
u( t) = e( ) d
Kp
Ti ed
Ti 0

e Si e=cte. e Kp e
t t

Ti = 1 repetition
The integral action will
Kp Kp
equate the proportional one
Ti ed = Ti
et = K p e t = Ti
in Ti time units if e is
constant (one repetition)
20
 Derivative action

w e u
+
- Kp Ampl.

e = w-y de
K pTd M
dt

The derivative term will smooth sharp changes in the

control signal due to fast changes in the error
21
 e=w-y
Derivative action
w y w
y
e( t ) = w ( t ) y ( t )
1 de
u ( t ) = K p e( t ) + e()d + Td
t Ti dt
t
u u

t
A P controller tuned with
high gain in order to get a
fast process response can
generate too strong u
changes and oscillations
t
If e decreases very fast, the
derivative term will decrease u,
avoiding oscillations
22
 Derivative action

de
u ( t) = K p Td
dt
Kp e

e Si e= a t e Kp Td a
t t

Td
After Td time units,
de
derivative and K p Td = K p Td a = K p at t = Td
dt
proportional terms will be
equal if e= a.t.
23
 Derivative action

PI PID

24
 Derivative action

w
y y
w

t t
de
u u ( t ) = K p Td u
dt
t t

Sharp changes in w Noisy process signals lead

cause big changes in u at to fast changing control
the time of change actions u 25
 Ideal PID (non interactive)

t
1 de( t )
u ( t ) = K p (e( t ) + e()d + Td ) e(t) = w(t) - y(t)
Ti 0 dt

I
+ u
e
w + P
-
y D

It is not physically implementable

Very sensitive to noises
Real zeros for Ti > 4Td 26
 Real PID (non interactive)

1
t
de
u ( t ) = K p e( t ) + e( )d + Td f
Ti 0 dt
Td def 1
+ ef = e( t ) filter in the error Ef ( s ) = E( s )
N dt Td
s +1
N
1 sTd
U(s) = K p 1 + + E( s )
Tsi 1 + sTd N

Physically implementable
Incorporates a filter in the derivative term
At high frequencies the maximum gain of the D term is Kp N
N : Maximum derivative gain. Typically N=10.
27
 Effect of Filters

without with

28
 PID (derivative action on y)
t
1 dy ( t)
u ( t) = K p ( e( t) + e( ) d Td f ) Honeywell type B
Ti 0 dt
dy f ( t)
0.1Td + y f ( t) = y( t)
dt
I
+ u
e
w + P
-
y -D

Used in the DCS

It avoids sharp changes in u when a step change is
given to w e = w -y 29
 PID modified proportional action

1
t
d yf
u ( t ) = K p (w ( t ) y( t )) +
Ti 0 e( ) d Td d t
The factor allows having a certain independence when tuning
the controller against load or set point changes

I
+ u
e
w I + Kp
-
y -D-1

30
 PID modified proportional action
1
t
dy f
with = 0 u ( t ) = K p ( y( t )) + e()d Td
Ti 0 d t

Honeywell type C

+ u
e
w I + Kp
-
y -D-1

31
 Series or Interactive PID

1 1 + Tdss
U(s) = K ps (1 + )( ) E (s)
Tiss 1 + 01. Tdss

I
w
+ e u
+ + P
-
y D

32
 Series or Interactive PID

1 1 + Tdss
U(s) = K ps (1 + )( ) E (s)
Tiss 1 + 01. Tdss

Used in the old analog or loop controllers

Equivalence tables between the parameters of series
and parallel PID types

F=1+Tds/Tis Kp= Kps F; Ti= Tis F; Td = Tds / F

Fs =0.5+(0.25-Td /Ti)0.5 Kps= Kp Fs; Tis = Ti Fs; Td s = Td / Fs

33
 Non linear PID

The gain is modified, so that the action of the controller

is stronger when the error is big and very smooth or
zero when the error is small or there are noises, etc

1
t
dy f
u ( t ) = K pf ( e) e( t ) + e( )d Td
Ti 0 dt
f ( e) function of the error, e.g. :
f ( e) = + (1 )e with, for instance, = 0.1

34
 Non linear PID

f(e) Non linear function of the error

Dead zone around e=0
High gain for big |e|
e
F(e) u
PID

There are no changes in u when e is small, (e.g. noises)

Increases the control actions if e is big
35
 Saturation in the instruments

4-20 mA

All actuators and transmitters have a limited range of

operation, with its signals been constrained to it (0 - 100 %)

36
 Reset wind-up

PID Process
t
1
u ( t ) = K p (e( t ) + e()d)
Ti 0

Delay in the actuation of the controller output that appears

when the value of the integral term exceed the allowable
range of the manipulated variable.
The implementation of the so called anti wind-up systems,
avoid the appearance of this phenomenon.

37
 Reset wind-up
Due to the
integral term PID Process

100%
t
1
u u ( t ) = K p (e( t ) + e()d)
Ti 0

without anti reset

w wind-up

y with anti reset

wind-up

38
 Anti-reset wind up
t
1 Key action: Stop the
1. u ( t ) = K p (e( t ) + e()d) integration if the integral
Ti 0
term exceeds the output
range Actuator range
e v u
Kp +

- +
Kp/Ti + 1/s

2. Tracking: If v >u, ew
then ew corrects the 1/ Tt
integral term until v = u Tt traking time constant
39
Anti-reset windup

Kp=4
Ti=1

40
 auto/man transfers

+
auto u
e
w PID
-
y manual

In a auto/man mode transfer u can suffer from strong changes

The controller should operate with smooth auto/man and
man/auto transfers (bumpless)
Changing the value of a parameter should be made without
strong output changes 41
Man follows the
Bumpless transfers
auto value in +
1/ Tr
automatic mode -
Auto/man

1/s +

Manual
PD man
Auto/man
auto u
Kp/Ti + 1/s +
e
Auto follows - +
the value man
in manual mode 1/ Tr
ew 42
 Control Pyramid

PID loops with Operation and control

good performance Economic in a process factory
optimization
are very important are organized in a set
for a proper of interacting layers
Advanced
process operation control MPC that provide different
any advanced functionalities
control or PID control / DCS
optimization
implementation
Field Instrumentation

Process

43
 What is good performance?

v DV y(t)
w u y
Controller Process
SP MV CV

Steady state errors

Settling time
Overshoot u(t)

Smooth control signal

Robustness
Output variance ..
 When using PID control?
PID controllers work well with most of the single input
single output (SISO) control problems (flow, pressure,
speed, ...)
Nevertheless, the PID may not be a good option when
dealing with difficult dynamics or very demanding
specifications:
Significant delay unstable systems
Non minimum phase multivariable interaction

u t t
45
 PID tuning
Dynamic behaviour of the
control loop depends on the
values of the PID parameters
Selection of the PID
parameters in order to obtain
a desired closed loop
behaviour
Kp, Ti, Td
Other parameters: N,Tr, ,
T,, constraints, ...
Several methods + process
knowledge
 Tuning criteria

Select the type of controller P, PI, PID, PD, type B,

C.. or other controller (DMC, IMC,...)
Tuning respect to set point or disturbance changes
(w or v)
Different control aims
Do not forget the manipulated variable
Robustness against changes in the process or the
operating point.
Compromise performance /robustness

47
 Controller types

PID is the right choice in slow processes without

a significant noise, such as temperature,
concentration and, in some cases pressure.
PI is the preferred choice most of the times
P is used in processes with an integrator o where a
zero steady state error is not important (e.g.
internal loops in cascades).
If the process have a significant delay use a Smith
Predictor. Use MPC in multivariable, constraint or
economic important process units.
48
 Tuning: SP or disturbances?
v
w u y
+ R G
-
Proceso

GR 1
y= w+ v
1 + GR 1 + GR
If the PID is tuned to obtain a good response against
disturbances, then R is fixed and the dynamical response
with respect to SP is also fixed. And viceversa.
PID: a single degree of freedom
49
 Disturbance / SP

Change
in SP

Disturbance
50 52

50
 PID Tuning methods

Trial and error methods

Experiment based methods
Perform an experiment in order to estimate certain
dynamic characteristics of the process
Compute the tuning parameters using tables or formulas
as a function of the estimated dynamical characteristics
of the process
Model based analytical methods
Automatic tuning methods

51
 Trial and Error
w w
y y

1 Increase Kp 2 Increase Td
Start from low Kp, and without integral or
derivative actions
w
y Increase Kp until a nice CV shape response is
obtained without using excessive MV. Do not
consider the steady state error
Increase a bit Td and Kp in order to improve
the response
3 Decrease Ti
Decrease Ti until the steady state error is
Agresive tuning / Robustness cancelled in a sensible time 52
 PID Tuning

First known PID tuning rule proposal:

Callander, A. (1934) Preliminary notes in
automatic control, Imperial Chemical
Industries (ICI), England. Internal report.
But this work remains unknown until 2004,
and the Ziegler-Nichols rules (1942) are cited
often as the first ones.
More than 1700 published rules

53
 Ziegler-Nichols methods
Tuning criterion: damping against disturbances (QDR)
Empirically developed for series PID (1942)
Two methods: Open and closed loop
Can be applied when 0.15 < d / < 0.6 in monotonous
processes
Provide good starting values that can be fine tuned

y
1
1/4 w

54
Open and closed loop methods

Closed loop experiment

Controller Process

Open loop experiment

Process

55
Open loop Ziegler-Nichols tuning
table
Type Gain Kp Integral Derivative
time time
P / (K d)
PI 0.9 /(K d) 3.33 d
Series PID 1.2 /(K d) 2 d 0.5 d

K process gain , d delay , time constant

Ti and Td in the same units as d
Notice that Ti = 4 Td
When applied to digital controllers, increase d by half a sampling period
56
 Step test Identification
Tangent of maximum slope
y Resting value

y
t = y/u
d
ds
u Ke
u
s + 1
t

Adequate for Ziegler-Nichols 57

 Step test Identification

y
= 1.5 (t2 - t1)
0.632y d = t2 -
y
0.283y
t = y/u
t1 t2
ds
u Ke
u
s + 1
t

Adequate for noisy systems 58

Open loop Ziegler-Nichols tuning
table
Type Gain Kp Integral Derivative
time time
P / (K d)
PI 0.9 /(K d) 3.33 d
Series PID 1.2 /(K d) 2 d 0.5 d

K process gain , d delay , time constant Units

Ti and Td in the same units as d
Notice that Ti = 4 Td
When applied to digital controllers, increase d by half a sampling period
59
 Two options for units
Kp % / %
w u
e mA
+
- R Process
%
Ing. % Ing.

% mA

w e u mA
+
- R Process
Ing. %
Ing.
Ing. mA
Kp % / Ing.
60
 Integral of the error
minimization
d
w e u Ke ds y
PID +
s + 1

y
min
K p , Ti , Td
e( t) dt MIAE

dt MISE
2
min e ( t )
K p , Ti , Td
w
min
K p , Ti , Td
e( t) tdt MITAE
error = f ( Kp , Ti, Td )
61
 Lopez et al. tuning table

Developed for Non interactive (parallel) PID (1967)

b
For disturbance rejection
d
K K = a
Tuning criteria: p

Integral of the error minimization:

b
d
MIAE |e| = a
MISE e2 Ti
MITAE |e|t b
Td
d
Based on First order plus delay model = a

The tables provide the a and b parameters of
the formulas
Can be applied to monotonous processes with
0.1 < d / < 1
62
 Lopez et al. tuning table
Parallel PID controllers
b
d
Criteria Proportional Integral Derivative K p K = a

MIAE a=1.435 a=0.878 a=0.482

b
b=-0.921 b=-0.749 b=1.137 d
= a
MISE a=1.495 a=1.101 a=0.560 Ti
b=-0.945 b=-0.771 b=1.006
b
MITAE a=1.357 a=0.842 a=0.381 Td d
b=-0.947 b=-0.738 b=0.995 = a

K in the same units as Kp
Disturbance rejection tuning
Can be used with monotonous processes with 0.1 < d / < 1
When applied to digital controllers, increase d by half a sampling period 63
Integral of the error minimization

w e u Ke ds y
PID +
s + 1

y
w
min
K p , Ti , Td
e( t) dt MIAE

dt MISE
2
min e ( t )
K p , Ti , Td

min
K p , Ti , Td
e( t) tdt MITAE
error = f ( Kp , Ti, Td )
64
 Rovira et al. tuning table
Parallel PI
Criteria Proportional Integral Derivative
b
d
MIAE a=0.758 a=-0.323 K p K = a
b=-0.861 b=1.020
MITAE a=0.586 a=-0.165 d
b=-0.916 b=1.030 = a + b
Parallel PID Ti
b
Td d
MIAE a=1.086 a=-0.130 a=0.348 = a
b=-0.869 b=0.740 b=0.914
MITAE a=0.965 a=-0.147 a=0.308
b=-0.855 b=0.796 b=0.929
K in the same units as Kp
Set point following tuning
Can be used with monotonous processes with 0.1 < d / < 1
When applied to digital controllers, increase d by half a sampling period 65
 Tuning

Type Kp Ti Desired closed

loop time
PI processes 2 + d
with integrator k ( + d )2 2 + d constant
PI 4 + d d Settling time/3
+
4 K 4

w y w y
1
s + 1

Lambda Tuning refers to all tuning methods where the control loop speed of
response is a selectable tuning parameter known as Lambda. Some rules recommend
values of higher than the open loop time constant 66
 Rivera-Morari IMC
Type Kp Ti Td recommended
>0.2 always
PI
> 1.7
K( + d ) d
Improved 2 + d +
d
PI > 1.7
2K 2 d
Parallel PID with 2 + d d d
+ > 0.25
filter 2 K ( + d ) 2 2 + d d

w 1 y Desired closed
s + 1 loop time constant

Practical = max (0.1, 0.8d) conservative: max (0.5, 4d)

67
 Model:
0.46e 0.87 s
0.96s + 1

K p = 1.7,
Ti = 1.17

Lambda tuning
= 1.5
68
 Direct synthesis
w u y
+
- R G

GR
Y(s) = W (s) M(s) = Desired closed
1 + GR loop TF

GR M (s)
M (s) = R (s) =
1 + GR G (s)(1 M (s))

69
 Automatic tuning methods

Most of the commercial controllers incorporate some methods

for automatic tuning (most of them autotuning)
Only in a few cases we find real adaptive control
Autotuning: The tuning procedure starts under operator demand

Step response
Relays method
Closed loop response
identification (Exact)
IFT
Gain scheduling

70
 Tuning in DCS

There are
applications
to help in the
automatic or
manual tuning
in the DCS

71
 Step response

When the autotuning function is activated, the controller is

switched into manual mode, then, it generates a step in
order to identify a first order plus delay model from which
the controller parameters are obtained using tuning tables.

e u
PID Process
SIPART (Siemens)
Pre-tuning:
EXACT, Electromax

72
 The relay method

When the autotuning function is activated, there is a

switching from the PID to a relay controller that creates
controlled oscillations in the process which are used to
identify some of its dynamic characteristics

e u
PID Process

ECA40 (Satt)
DPR9000 (Fisher)
Astrom, Hagglund 1984

73
 The relay method

T oscillation period
A amplitude of the first harmonic
y
T
w
A
t d
e
Process
-d u
u

t
Other options: relay with hysteresis or
additional loops are added in order to force
Advantage: The process is under
the generation of oscillations 74
controlled oscillations
 The Exact method
EXact Adaptive Controller Tuning (Foxboro)
Continuous closed loop tuning
If the error exceeds a range, then a process identification
procedure based on pattern recognition is started
The controller computes the new tuning in real time using
modified Ziegler-Nichols tables plus some rules
The desired dynamics is specified in terms of overshoot and
damping
e
PID Process
Pretuning
with the step
method 75
 Exact activation Change in w
e
y

w NB

y Wmax Disturbance

w
The procedure is activated automatically if the error is outside
the error band NB and the second pick appears before Wmax
sg. after the first one
If no second pick appears before Wmax, the process is
considered a overdamped one 76
Iterative Feedback Tuning IFT
Closed loop tuning
Two special experiments
are performed in order to
estimate the gradient of a
quadratic cost function
with respect to the PID
parameters
Tuned by cost
minimization

77
 Gain scheduling

Adjustment
table
New tuning
w
PID Process

The controller parameters are adjusted using a pre-

computed table function of some operating condition:
e.g. the set point value. It adapts the controller to the
new process dynamics
78
 Kp Gain Scheduling
-1.32

-0.5

130 C 145 C
SP
79
 Systems with delay

w u y
+
- R Ge-ds

If the delay is higher than the process time constant, the system
is difficult to tune.
The Smith predictor is a controller that improves the time
response of this type of processes. It needs to know the model
Ge-ds
80
 Delays: Smith Predictor

w u y
R Ge-ds

Gm(1-e-ds)

y = Ge ds u = Ge ds R [w y G m (1 e ds )u ] =
= Ge ds R [w Ge ds u G m (1 e ds )u ]
si G = G m y = Ge ds R [w Gu ]
81
 Smith Predictor

y =e ds GR [w Gu ]

y
w u
R G e-ds

Equivalent diagram

R can be tuned as if there were no delay

82
 0.46e 0.87 s
0.96s + 1 Smith Predictor
K p = 1.32, Ti = 0.96

PI Smith

83
 Controller monitoring
A typical factory has hundred Temp

/thousands of control loops in TC

operation FC TC TT
TT
Reactante

Automatic supervision methods Ti FT Tr T

Reactor
FC FT

are required for these task. They q Refrigerante

should be able to work, both with LT

AT AC Comp.

isolated loops or with control LC Producto

structures (cascades, etc.)

Automatic controller monitoring is
feasible because of the availability
of data, DCS systems, computers
and algorithms.

84
 Controller monitoring

Focuses the attention in determining if a control loop

behaves properly and, if not, if the cause of the bad
behaviour is due to the controller

w u y
Controlador Proceso

85
 Model based methods
v

w u y
Controlador Proceso
They require the
identification of the process
either in open or closed
loop
Then, it is possible to
analyse the dynamic
characteristics of the Model
process (or the closed loop)
and decide about the state
and performance of the
loop. Analysis
86
 Methods based on normal operating
data
Harris index
Harris (1989)
Desborough and Harris (1992, 1994)
Similar ones: Kobuz and Garcia CLP(1993), Jofriet (1996), Tyler and
Morari (1995),etc.
Measures the gap between the control loop performance and the best
attainable one (in terms of minimum output variance)
Predictability Index
Thornhill (1999), Ghraizi, Prada and Martinez (2003)
Measures the closed loop performance based on the error predictability
They are complementary

87
 Harris index

It compares the behaviour of the controller with the best

possible one (minimum variance)
Performance is measured in terms of output variance
It does not measure if the behaviour of the loop is good or bad,
but the distance to the best possible controller.
Can be estimated from normal operating data, but requires the
knowledge of the process delay.

y w 2mv ( t )
HI( t ) = 2
e (t )
HighVariance Low Variance It goes from 0 to 1
Bad to good
88
 Flow control. Slave in a cascade.
Coeficientes del model de Harris

Coef icientes del modelo de Harris

0.8

0.6

0.4
CAUDAL DE OVERFLAS D-36 27FC0207
0.2
50
variable controlada PV
40 Consigna 0

30 -0.2
20 40 60 80 100 120
20

10 Indice de Harris

0 0.55
2000 4000 6000 8000 10000 12000 14000 16000
0.5

Variable manipulad OP 0.45

100 0.4
variable manipulada OP
0.35

0.3

50 0.25

0.2
HI
0.15
0 0.1
2000 4000 6000 8000 10000 12000 14000 16000
0.05

2 4 6 8 10 12 14 16

89
 Error Predictability
e(t) = w(t) y(t)
e(t) predictible

time

t t+b Non-
predictible

After some time, b, the error signal of a control loop that

behaves properly must be only the result of unpredictable
stochastic disturbances. Hence their value should not be
predicted well using information up to time instant t only.
b: bigger than the settling time of a well performing loop
90
 Bad performance:
Error Predictability
Signal patterns that
e(t) can be identified
and predicted after
Good t+b
time
t t+b
e(t) Oscilations

time
t
e(t)
Steady state error

time
t
91
 Error residual analysis
Error residuals: Difference between the error at time instant
t+b and the error prediction at t+b computed from data up to
time instant t
e(t)
e( t + b | t )

t
e( t + b )

r ( t ) = e( t + b) e( t + b | t ) e( t ) = w ( t ) y ( t )
92
 Performance index PI

Performance index is computed in a similar way to

the Harris one:
Variance of the
(t) 2
residuals of the error
PI( t ) = 1 r
(t) 2
e Variance of the
Variance estimation error
1 N 1 N
( ) ( )
2 2 2 2
r ( t ) r e ( t ) e
N 1 t =1 N 1 t =1
r e

Does not require process delay knowledge

Uses data from normal operation
93
 Industrial Data/Cascaded Flow loop

MV saturation due to inadequate 17 batches of 1000 data each b= 12,

SP fixed by the master m= 30, tm = 5

94
Flow loop / error predictions

The PI trend indicates a local problem, not bad tuning

95
 Pressure loop

Bad tuning, even if there is a good SP

following

Error predictions
96
 Control structures
Split-range

Cascade
Feedforward

ratio

Override 97
Thank you for your attention

98