pid控制外文加中文文献可编辑.docx
《pid控制外文加中文文献可编辑.docx》由会员分享,可在线阅读,更多相关《pid控制外文加中文文献可编辑.docx(16页珍藏版)》请在冰豆网上搜索。
![pid控制外文加中文文献可编辑.docx](https://file1.bdocx.com/fileroot1/2022-11/27/5de8316a-581f-4e6b-aa8a-ab37ad57ca41/5de8316a-581f-4e6b-aa8a-ab37ad57ca411.gif)
pid控制外文加中文文献可编辑
pid控制外文加中文文献(可编辑)
PIDcontroller
FromWikipediathefreeencyclopedia
Aproportional–integral–derivativecontrollerPIDcontroller
isagenericcontrolloopfeedbackmechanismwidelyusedinindustrialcontrolsystemsAPIDcontrollerattemptstocorrecttheerrorbetweenameasuredprocessvariableandadesiredsetpointbycalculatingandthenoutputtingacorrectiveactionthatcanadjusttheprocessaccordingly
ThePIDcontrollercalculationalgorithminvolvesthreeseparateparameterstheProportionaltheIntegralandDerivativevaluesTheProportionalvaluedeterminesthereactiontothecurrenterrortheIntegraldeterminesthereactionbasedonthesumofrecenterrorsandtheDerivativedeterminesthereactiontotherateatwhichtheerrorhasbeenchangingTheweightedsumofthesethreeactionsisusedtoadjusttheprocessviaacontrolelementsuchasthepositionofacontrolvalveorthepowersupplyofaheatingelementBy"tuning"thethreeconstantsinthePIDcontrolleralgorithmthePIDcanprovidecontrolactiondesignedforspecificprocessrequirementsTheresponseofthecontrollercanbedescribedintermsoftheresponsivenessofthecontrollertoanerrorthedegreetowhichthecontrollerovershootsthesetpointandthedegreeofsystemoscillationNotethattheuseofthePIDalgorithmforcontroldoesnotguaranteeoptimalcontrolofthesystemorsystem
stability
SomeapplicationsmayrequireusingonlyoneortwomodestoprovidetheappropriatesystemcontrolThisisachievedbysettingthegainofundesiredcontroloutputstozeroAPIDcontrollerwillbecalledaPIPDPorIcontrollerintheabsenceoftherespectivecontrolactionsPIcontrollersareparticularlycommonsincederivativeactionisverysensitivetomeasurementnoiseandtheabsenceofanintegralvaluemaypreventthesystemfromreachingitstargetvalueduetothecontrolaction
AblockdiagramofaPIDcontroller
NoteDuetothediversityofthefieldofcontroltheoryandapplicationmanynamingconventionsfortherelevantvariablesareincommonuse
1Controlloopbasics
AfamiliarexampleofacontrolloopistheactiontakentokeeponesshowerwaterattheidealtemperaturewhichtypicallyinvolvesthemixingoftwoprocessstreamscoldandhotwaterThepersonfeelsthewatertoestimateitstemperatureBasedonthismeasurementtheyperformacontrolactionusethecoldwatertaptoadjusttheprocessThepersonwouldrepeatthisinput-outputcontrolloopadjustingthehotwaterflowuntiltheprocesstemperaturestabilizedatthedesiredvalue
FeelingthewatertemperatureistakingameasurementoftheprocessvalueorprocessvariablePVThedesiredtemperatureiscalledthe
setpointSPTheoutputfromthecontrollerandinputtotheprocessthetappositioniscalledthemanipulatedvariableMVThedifferencebetweenthemeasurementandthesetpointistheerroretoohotortoocoldandbyhowmuchAsacontrolleronedecidesroughlyhowmuchtochangethetappositionMVafteronedeterminesthetemperaturePVandthereforetheerrorThisfirstestimateistheequivalentoftheproportionalactionofaPIDcontrollerTheintegralactionofaPIDcontrollercanbethoughtofasgraduallyadjustingthetemperaturewhenitisalmostrightDerivativeactioncanbethoughtofasnoticingthewatertemperatureisgettinghotterorcolderandhowfastandtakingthatintoaccountwhendecidinghowtoadjustthetapMakingachangethatistoolargewhentheerrorissmallisequivalenttoahighgaincontrollerandwillleadtoovershootIfthecontrollerweretorepeatedlymakechangesthatweretoolargeandrepeatedlyovershootthetargetthiscontrolloopwouldbetermedunstableandtheoutputwouldoscillatearoundthesetpointineitheraconstantgrowingordecayingsinusoidAhumanwouldnotdothisbecauseweareadaptivecontrollerslearningfromtheprocesshistorybutPIDcontrollersdonothavetheabilitytolearnandmustbesetupcorrectlySelectingthecorrectgainsforeffectivecontrolisknownastuningthecontroller
IfacontrollerstartsfromastablestateatzeroerrorPVSPthenfurtherchangesbythecontrollerwillbeinresponsetochangesin
othermeasuredorunmeasuredinputstotheprocessthatimpactontheprocessandhenceonthePVVariablesthatimpactontheprocessotherthantheMVareknownasdisturbancesandgenerallycontrollersareusedtorejectdisturbancesandorimplementsetpointchangesChangesinfeedwatertemperatureconstituteadisturbancetotheshowerprocess
IntheoryacontrollercanbeusedtocontrolanyprocesswhichhasameasurableoutputPVaknownidealvalueforthatoutputSPandaninputtotheprocessMVthatwillaffecttherelevantPVControllersareusedinindustrytoregulatetemperaturepressureflowratechemicalcompositionspeedandpracticallyeveryothervariableforwhichameasurementexistsAutomobilecruisecontrolisanexampleofaprocesswhichutilizesautomatedcontrol
DuetotheirlonghistorysimplicitywellgroundedtheoryandsimplesetupandmaintenancerequirementsPIDcontrollersarethecontrollersofchoiceformanyoftheseapplications
2PIDcontrollertheory
NoteThissectiondescribestheidealparallelornon-interactingformofthePIDcontrollerForotherformspleaseseetheSection"AlternativenotationandPIDforms"
ThePIDcontrolschemeisnamedafteritsthreecorrectingtermswhosesumconstitutesthemanipulatedvariableMVHence
wherePoutIoutandDoutarethecontributionstotheoutputfromthe
PIDcontrollerfromeachofthethreetermsasdefinedbelow
21Proportionalterm
TheproportionaltermmakesachangetotheoutputthatisproportionaltothecurrenterrorvalueTheproportionalresponsecanbeadjustedbymultiplyingtheerrorbyaconstantKpcalledtheproportionalgain
Theproportionaltermisgivenby
Where
PoutProportionaloutput
KpProportionalGainatuningparameter
eErrorSPPV
tTimeorinstantaneoustimethepresent
ChangeofresponseforvaryingKpAhighproportionalgainresultsinalargechangeintheoutputforagivenchangeintheerrorIftheproportionalgainistoohighthesystemcanbecomeunstableSeethesectiononLoopTuningIncontrastasmallgainresultsinasmalloutputresponsetoalargeinputerrorandalessresponsiveorsensitivecontrollerIftheproportionalgainistoolowthecontrolactionmaybetoosmallwhenrespondingtosystemdisturbances
IntheabsenceofdisturbancespureproportionalcontrolwillnotsettleatitstargetvaluebutwillretainasteadystateerrorthatisafunctionoftheproportionalgainandtheprocessgainDespitethe
steady-stateoffsetbothtuningtheoryandindustrialpracticeindicatethatitistheproportionaltermthatshouldcontributethebulkoftheoutputchange
comgralterm
ThecontributionfromtheintegraltermisproportionaltoboththemagnitudeoftheerrorandthedurationoftheerrorSummingtheinstantaneouserrorovertimeintegratingtheerrorgivestheaccumulatedoffsetthatshouldhavebeencorrectedpreviouslyTheaccumulatederroristhenmultipliedbytheintegralgainandaddedtothecontrolleroutputThemagnitudeofthecontributionoftheintegraltermtotheoverallcontrolactionisdeterminedbytheintegralgainKi
Theintegraltermisgivenby
ChangeofresponseforvaryingKiWhere
IoutIntegraloutput
KiIntegralGainatuningparameter
eErrorSPPV
τTimeinthepastcontributingtotheintegralresponse
Theintegraltermwhenaddedtotheproportionaltermacceleratesthemovementoftheprocesstowardssetpointandeliminatestheresidualsteady-stateerrorthatoccurswithaproportionalonlycontrollerHoweversincetheintegraltermisrespondingtoaccumulatederrorsfromthepastitcancausethepresentvaluetoovershootthesetpointvalue
crossoverthesetpointandthencreateadeviationintheotherdirectionForfurthernotesregardingintegralgaintuningandcontrollerstabilityseethesectiononlooptuning
23Derivativeterm
TherateofchangeoftheprocesserroriscalculatedbydeterminingtheslopeoftheerrorovertimeieitsfirstderivativewithrespecttotimeandmultiplyingthisrateofchangebythederivativegainKdThemagnitudeofthecontributionofthederivativetermtotheoverallcontrolactionistermedthederivativegainKd
Thederivativetermisgivenby
ChangeofresponseforvaryingKdWhere
DoutDerivativeoutput
KdDerivativeGainatuningparameter
eErrorSPPV
tTimeorinstantaneoustimethepresent
ThederivativetermslowstherateofchangeofthecontrolleroutputandthiseffectismostnoticeableclosetothecontrollersetpointHencederivativecontrolisusedtoreducethemagnitudeoftheovershootproducedbytheintegralcomponentandimprovethecombinedcontroller-processstabilityHoweverdifferentiationofasignalamplifiesnoiseandthusthisterminthecontrollerishighlysensitivetonoiseintheerrortermandcancauseaprocesstobecomeunstableif
thenoiseandthederivativegainaresufficientlylarge
24Summary
TheoutputfromthethreetermstheproportionaltheintegralandthederivativetermsaresummedtocalculatetheoutputofthePIDcontrollerDefiningutasthecontrolleroutputthefinalformofthePIDalgorithmis
andthetuningparametersare
KpProportionalGain-LargerKptypicallymeansfasterresponsesincethe
largertheerrorthelargertheProportionaltermcompensationAnexcessivelylargeproportionalgainwillleadtoprocessinstabilityandoscillation
KiIntegralGain-LargerKiimpliessteadystateerrorsareeliminatedquickerThetrade-offislargerovershootanynegativeerrorintegratedduringtransientresponse