Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
0Activity
0 of .
Results for:
No results containing your search query
P. 1
Pid Explained

Pid Explained

Ratings: (0)|Views: 3|Likes:
Published by mkhalil410

More info:

Published by: mkhalil410 on Oct 12, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

10/12/2013

pdf

text

original

 
P
ID controllers - named after the Proportional, Integral andDerivative control actions they perform - are used in thevast majority of automatic process control applications inindustry today. PID controllers are responsible for regulatingflow, temperature, pressure, level, and a host of other industrial process variables. This tutorial reviews the application of PIDcontrollers, explains the P, I and D control modes and units, andhighlights the three controller structures used in industrialcontrollers.Withoutautomatic controllers,allregulationtaskswillhavetobedone manually. For example: To keep constant the temperatureof water discharged from an industrial gas-fired heater, anoperator has to watch a temperature gauge and adjust a gascontrol valve accordingly (Figure 1). If the water temperature becomes too high, the operator has to close the gas control valvea bit - just enough to bring the temperature back to the desiredvalue. If the water becomes too cold, he has to open the valveagain.To relieve our operator from the tedious task of manual control,we automate the controls - i.e. we install a PID controller (Figure2). The controller has a Set Point (SP) that the operator canadjust to the desired temperature. We also have to automate thecontrol valve by installing an actuator (and perhaps a positioner)so that the Controller's Output (CO) can change the valve's position. And finally, we'll provide the controller with anindication of the temperature or Process Variable (PV) byinstalling a temperature transmitter. The PV and CO are mostlytransmitted via4-20mAsignals.So, when everything is up and running, our PID controllecompares the process variable to its set point and then calculatesthe differencebetweenthe twosignals,alsocalledtheError(E).As we saidabove, a PID controller has proportional, integral andderivative control modes. These modes each react differently tothe error, and also, the degree of control action is adjustable for eachmode.The proportional control mode changes the controller output in proportion to the error (Figure 3). The adjustable setting here iscalled the Controller Gain (Kc), sometimes also referred to as aPIDcontroller's P-settingorits proportionalsetting.The control action is proportional to both the controller gain andthe error. A higher controller gain will increase the amount of outputactionandsowillalargererror.Although most controllers use controller gain (Kc) as the proportionalsetting,somecontrollers useProportionalBand
Figure1. ManualControlFigure 3. ProportionalControlAction
Then,basedon the error, a fewadjustable settingsanditsinternalstructure (described below), the controller calculates an outputthat positions the control valve. If the actual temperature isabove its set point, the controller will reduce the valve positionandvice versa.l
Figure 2. AutomaticContro
ProportionalControl
ManualcontrolAutomaticcontrolP,IandDControlModes
PIDControllersExplained
by- Plant Automation Services, Inc.
Jacques Smuts, Ph.D.
Reprinted with permission from , February 2002
Canadian Process Equipment & Control News
Transmitter 
Control Valve
Controller 
ProcessPVCOSP
P
Temperature GaugeHand ValveProcessOperator 
 
(PB), which is expressed in percent. Table 1 shows therelationshipbetween Kcand PB.0.1 10000.2 5000.5 2001 1002 505 2010 10The use of proportional control alone has a large drawback -Offset. Offset is a sustained error that cannot be eliminated by proportional control alone. For example, let's consider controlling the water level in the tank in Figure 4 with a proportional-onlycontroller.As long as the flow out of the tank remains constant, the level(which is our process variable in this case) will remain at its set point.But, if the operator should increase the flow out of the tank, thetank level will begin to decrease due to the imbalance betweeninflow and outflow. While the tank level decreases the error increases and our proportional controller increases the controller output proportional to this error (Figure 5). Consequently, thevalve controlling the flow into the tank opens wider and morewater flowsintothe tank.As the level continues to decrease, the valve continues to openuntil it gets to a point where the inflow matches the outflow. Atthis point the tank level remains constant, and so does the error.Then, because the error remains constant our P-controller willkeep its output constant and the control valve will hold its position. The system now remains at balance with the tank levelremainingbelowitssetpoint. This residual erroris calledOffset.With our P-controller the offset will remain until the operator manually applies a bias to the controller's output to remove theoffset. It is said that the operator has to manually "Reset" thecontroller. Or…wecan addIntegralactiontoourcontroller.The concept of manual reset as described above led to thedevelopmentofautomaticresetorIntegralControl,asweknow ittoday. The integral control mode of a controllerproduces a long-term corrective change in controller output, driving the error or offsettozero.Integralactionappearsasaramp ofwhichtheslopeisdetermined by the size of the error, the controller gain and the Integral Time(Ti),also calledtheI-settingofthecontroller(Figure6).Most controllers use integral time in minutes as the unit fointegral control, but some others use integral time in seconds,Integral Gain in Repeats / Minute or Repeats / Second. Table 2compares thedifferentintegralunits.IntegralTime IntegralGain0.05 3 20 0.3330.1 6 10 0.1670.2 12 5 0.08330.5 30 2 0.03331 60 1 0.01672 120 0.5 0.008335 300 0.2 0.0033310 600 0.1 0.0016720 1200 0.05 0.00083Integral control eliminates offset. Figure 7 shows the same levelcontrol setupasbefore,but thistime we haveuseda PI controller.A PI controller simply adds together the output of the P and Imodes of the controller. The integral action raises the controller outputfar enoughtobringthelevelback toitssetpoint.ControllerGain(Kc) ProportionalBand(PB) %Minutes Seconds Rep/Minute Rep/Second
Table 1. RelationshipbetweenKc andPBFigure4. LevelcontrolFigure5. P-ControlFigure6. IntegralControlActionTable2. IntegralControlUnits
IntegralControl
PVCOSP
 
50.049.8
P
c
 K  PB
%100
=
∫ 
×=
dt  E  K  Action I 
ic
1
Reprinted with permission from , February 2002
Canadian Process Equipment & Control News
 
Figure 7. PI-ControlFigure8. DerivativeControlActionFigure 9. PID controlFigure10. Seriescontrollerstructure.Figure11. Idealcontrollerstructure.
DerivativeControlSeriesIdealParallel
The third control action in a PID controller is derivative.Derivative control is rarely used in controllers. It is verysensitive to measurement noise and it makes tuning very difficultif trial and error methods are applied. Nevertheless, derivativecontrol can make a control loop respond faster and with lessovershoot.The derivative control modeproduces an output based on the rateof change of the error (Figure 8). Derivative action is sometimescalled Rate. Its action is dependent on the rate of change (or slope) of the error. It has an adjustable setting called DerivativeTime(Td),whichistheD-settingofthecontroller.Two units are used for the derivative setting of a controller:minutesandseconds.Derivative control appears to have predictive or anticipativecapabilities. Technically this is not true, but PID control does provide more control action sooner than possible with P or PIcontrol. To see this, compare the initial controller response of Figure 9 (PID control) with that in Figure 7 (PI control). Alsonote how derivative control reduces the time it takes for the leveltoreturntoitsset point. Also notethatwithderivative control thecontroller output appears noisier. This is due to the derivativecontrol mode's sensitivitytomeasurementnoise.Controller manufacturers integrate the P, I and D-modes intothree different arrangements or controller structures. These arecalled Series, Ideal and Parallel controller structures. Somecontroller manufacturers allow you to choose between differentcontroller structures as a configuration option in the controller software.This very popular controller structure is also called the Classical,Real or Interacting structure. The original pneumatic andelectroniccontrollers hadthisstructure and we still findit inmostPLCs and DCSs today. Most of the controller tuning rules are basedonthiscontrollerstructure.Also called the Non-Interacting, Standard or ISA structure, thiscontroller structure was popularized with digital control systems.If no derivative isused(i.e.Td= 0),the seriesand ideal controller structuresbecomeidentical.Academic-type textbooks generally use the parallel form of PIDcontroller, but it is also used in some DCSs and PLCs. Thisstructure is simple to understand, but really difficult to tune. Thereason is that it has no controller gain, but has a proportional gaininstead. Tuning should be done by adjusting all the settingssimultaneously. Trynottousethis structureifpossible.
ControllerStructures
dt dE  K  Action D
c
×=+×+=
∫ 
dt dt  E  E  K CO
ic
11
++=
∫ 
dt dE dt  E  E  K CO
ic
1
∫ 
++×=
dt dE  K dt  E  K  E  K CO
i p
Reprinted with permission from , February 2002
Canadian Process Equipment & Control News

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->