CONTROL SYSTEM TECHNOLOGIES
1 - INTRODUCTION TO CONTROL SYSTEM
Industrial control systemsareessentialinmodernindustriesbecausethey
automate the monitoring and regulation of keyprocessvariablessuchas
temperature,pressure,flow,andlevel.Theprimaryaimofthesesystemsis
to ensurethattheconversionofrawmaterialsintoproductsoccurssafely,
efficiently, and with quality. By using sensors, actuators, controllers, and
feedback mechanisms, industrial control systems maintain processes at
desired operating points and quickly correct any deviations from set
targets.
Process control specifically deals with continuous processes, where feed
materials are transformed into finished products. Different types of
variables are measured in real time and automatically regulated to
maintain the process at its optimum state.Theadvantageisthatprocess
control systems reduce human intervention, minimize errors, andimprove
product consistency.
Beyond process industries, industrial control also applies to mechanical
systems, such as robots andmachinetools.Inthesecontexts,theprimary
variables controlled include position, velocity, and torque (or force).
Accurate control of these variables ensures that the mechanical system
operates smoothly, efficiently, and without damaging products or itself.
Feedback Control Loop
Afundamentalconceptinindustrialcontrolisthefeedbackcontrolloop.In
simpleterms,afeedbackcontrolloopcontinuouslymeasurestheoutputof
a process (the process variable
), compares it to the desired value (the
setpoint
),calculatestheerror(differencebetweendesiredandactual),and
uses this error to adjust the process. The effectiveness of such systems
1
dependsnotjustonthecontrolleritselfbutalsoonthequalityandspeedof
sensors and actuators, which mustprovidetimelyandaccuratedataand
responses.
Piping & Instrumentation (P&I) Diagrams
P&Idiagramsaretechnicalschematicsthatprovideadetailedmapofthe
components in a plant, including pipes, valves, sensors, and controllers,
and howthey’reinterconnected.Thesediagramsarecrucialforengineers,
as they reveal both the flow of materials and the arrangement of control
systems. Devices in a P&I diagram are labeled with specific codes—a
combination of letters and numbers—that quickly convey the type of
measurement(forexample,Tfortemperature),thefunction(suchasCfor
controllerorVforvalve),andtheloopnumber.Thisstandardizationmakes
it easier to understand and maintain complex systems.
Sensors (Transducers) and Their Characteristics
Sensors are the "eyes and ears" of industrial control systems. They detect
physicalorchemicalproperties—liketemperatureorpressure—andconvert
them into electrical signals that can be processed by controllers. Several
characteristics define a sensor’s suitability, such as
Accuracy is how close the sensor’s reading is, to the true value;
●
systematic errors can be minimized through calibration.
Precision refers to the repeatability of the sensor—whether it can
●
produce the same result under the same conditions (crucial for stability)
Rangeability (Turndown) is the span between the maximum and
●
minimum values the sensor can accurately measure.
RangeandSpanaretheminimum-to-maximumvalues(range)and
●
the difference between them (span).
CharacteristicCurveandScaleFactorindicatehowthesensoroutput
●
relates to the measured input; ideally, this is a linear relationship. 2
Sensitivity shows how much the output changesforagivenchange
●
in input; higher sensitivity means the sensor can detect small variations.
Resolution is the smallest detectable change by the sensor.
● Sensor Dynamics describes how quickly the sensor responds to
●
changes; slower sensors can introduce delays (lag) in the control system.
Types of Sensors in Process Industry
Different variables require different types of sensors:
Temperature Sensors, such as thermocouples, RTDs and thermistors
● Pressure Sensors: manometers and bourdon tubes
● Flow Sensors like Venturi tubes, magnetic, turbine, ultrasonic, and
●
coriolis flowmeters.
Level Sensors, such as float-actuated sensors, bubbler, differential
●
pressure, conductivity, capacitance, and ultrasonic sensors
Chemical Sensors that measure pH, viscosity, humidity, and more
● Sensors for Mechanical Systems which measure position, velocity
●
and force
Actuators
Actuators are the devices that actually make physical changes to the
process based on controller commands (i.e. a control valve adjusts the
flowratebychangingitsposition).Theperformanceofactuatorsiscritical
because they have physical limits:theycansaturate(notmovebeyonda
certain point), have rate limits (cannot move too fast) and introduce
nonlinearities (behavior changes at different positions or speeds).Control
systems must be designed to account for these real-world limitations to
avoid instability or poor process control.
Valves are a common type of actuator, used especially for flow control.
Their "characteristic" defines how flow changes as the valve opens: flow
3
increases in direct proportion to valve position, each equal step by a
consistent percentage.Thecorrectchoiceofvalvetypeandcharacteristic
isvitalforefficientandstableprocesscontrol.Also,asvalvesage,theycan
develop issues like friction and hysteresis.
Some critical issues in industrial control includenonlinearitiesandwearin
actuators, delays introduced by sensors, and the interplay between all
these components. If not properly addressed, these factors can degrade
system performance, leading to oscillations, slow responses, or even
instability.
2. PID CONTROL
PID controllers implement a Proportional-Integral-Derivativecontrollaw
and they are very successful for many reasons. Firstly they provide
satisfactory performances by keeping reduced costs, being an industrial
standard. They are usually sufficient: there is no reason to use more
4
complex control laws, considering that the best control system is the
simplest one that meets the requirements. Lastly, they are often used as
basesofmorecomplexcontrolschemes,obtainingasignificantincrement
in the performance with a reasonable increment of the complexity of the
design.
Analyzing in details the three mainactionofPIDcontrollers,thefirstoneis
the Proportional action, which has the following formula:
(
) =
(
)
+
whereKpistheproportionalgain, isthe‘reset’termthatcanbechosen
in order to guarantee a null steady state errorand isthecontrolerror
()
.
( = − )
Then, the Integral action is proportional to theintegralofthecontrolerror,
andit allowstheusertoachieveanullsteady-stateerrorbyselectingthe
correct value of :
1
(
)
=
∫ (τ)τ
0
If we combine together the integral and proportional actions we get a PI
controller, whose transfer function is: 1
()
=
(
1 + )
where T
istheintegraltimeconstant,whichparticipatesincalculatingthe
i
processgain ,anditcanbesetinordertooptimizethesystemstep
=
response. It’s important not to increase T excessively, because it can
i
converge to a proportional controller; also because the controller
|
| < 1
risks being too aggressive.
The Derivative action is proportional to the first derivative of the control
erroranditprovidesthepredictionofthecontrolerrorattimet+T (before
d
it occurs): ()
(
)
=
5
where T is the derivative time constant, and
d ()
( + )
≃
(
)
+
By increasing T the action is more aggressive and more sensitive to
d
changes, useful for fast systems but riskywhenwehaveasignalaffected
by noise (derivative tends to infinite).
Joining these three actions together we can obtain PID controllers, which
have different forms:
- Ideal form(non interactive) 1
() =
(
1 + +
)
Thisisthemostdirectform,oftenused
intheoryandacademictexts,because
it clearly separates the three actions:
proportional, integral, and derivative
- Series form 1
'() =
'
(
1 + )(1 +
'
)
In this case, the integral and
derivative actions influence
eachother(they“interact”),and
forthisreasonit’salsocalledthe
“interacting” form.
Remember that it is always possible to pass from the ideal to the serial
form,notviceversa(becauseofthesquarerootspresentintheconversion
formulae).
- Parallel form
''() = + +
6
Where the actions are perfectly independent: each term is regulated
separately.
In order to obtain a proper controller transfer function and to avoid the
amplification of the high-frequency noises, it’s necessary to filter the
derivative actionby using a first order system:
(
)
=
(
1 + )
(
)
1+
The best choice would be the crossover frequency,but is hard to choose
the best value of it, so it can be used as the cut-off frequencyofthe
filter. The right way to filter the noisewouldbetoapplythefilteronallthe
control actions (proportional, integral and derivative), rather then only on
the derivative one, and in this case we talk about the output-filtered PID
controller:
1
1
() =
(
1 + +
)
+1
Howeverthismethodisdifficulttoimplement,soweusethemethodshown
at the beginning ( ).
Whenastepisappliedasaset-pointsignal,thederivativeactionproduces
a signal with infinite amplitude, that is reflected in the control actionasa
peak called ‘
d
erivative kick
’. So as to avoid this problem (which would
y
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
Scarica il documento per vederlo tutto.
-
![System theory and avanced control - Appunti di Controllo nonlineare: regolazione (Set-Point Stabilization), equazioni di Solvibilità, linearizzazione, Gain Scheduling e controllo integrale]()
System theory and avanced control - Appunti di Controllo nonlineare: regolazione (Set-Point Stabilization), equazio…
-
![Appunti - Technologies for Informative System]()
Appunti - Technologies for Informative System
-
![Appunti di Industrial Technologies]()
Appunti di Industrial Technologies
-
![Appunti completi Web Design and Technologies]()
Appunti completi Web Design and Technologies
Accedi a tutti gli appunti
Tutor AI: studia meglio e in meno tempo
Appunto