spe-227896 通过融合物理模型与机器学习算法，提升自动滑动性能.pdfVIP

SPE-227896-MS

EnhancingAutomatedSlidingPerformanceThroughtheIntegrationof

PhysicsModelsandMachineLearningAlgorithms

M.Summersgill,J.Xu,A.Jamieson,T.Chen,S.Menand,M.Mahjoub,A.McGregor,J.Cantwell,andP.Landrio,

HelmerichPayne

Copyright2025,SocietyofPetroleumEngineersDOI10.2118/227896-MS

ThispaperwaspreparedforpresentationattheSPEAnnualTechnicalConferenceandExhibitionheldinHouston,Texas,USA,20-22October2025.

ThispaperwasselectedforpresentationbyanSPEprogramcommitteefollowingreviewofinformationcontainedinanabstractsubmittedbytheauthor(s).Contents

ofthepaperhavenotbeenreviewedbytheSocietyofPetroleumEngineersandaresubjecttocorrectionbytheauthor(s).Thematerialdoesnotnecessarilyreflect

anypositionoftheSocietyofPetroleumEngineers,itsofficers,ormembers.Electronicreproduction,distribution,orstorageofanypartofthispaperwithoutthewritten

consentoftheSocietyofPetroleumEngineersisprohibited.Permissiontoreproduceinprintisrestrictedtoanabstractofnotmorethan300words;illustrationsmay

notbecopied.TheabstractmustcontainconspicuousacknowledgmentofSPEcopyright.

Abstract

Automatedslidingsolutionshaveexistedintheindustryforyears.Inearlydeployments,simplymeeting

steeringrequirementswithnear100%automationevenatthecostofdrillingperformancewasconsidered

atechnicalsuccess.Now,algorithmsareexpectedtomeetorexceedperformancestandardssetbythebest

drillers.Amulti-facetedapproachleveragingpreciserigcontrol,physicsmodels,andmachinelearning

techniquesaimstodeliverconsistent,best-in-classperformanceinascalablemanner.

Thedynamicsofslidingwasmodeledusinghighfidelitytime-domainfiniteelementmodelacrossa

broadrangeofoperatingconditions.Areducedordermodel,basedonasetofkeyparameters,wasfit

totheresultingdatasettoeffic