第28卷第1期2023年2月工业工程与管理IndustrialEngineeringandManagementVol.28No.1Feb.2023基于DT的售后设备预测性维护协同模式研究李晓,陈雨晨*,阮渊鹏,王雷,劳晓云(杭州电子科技大学管理学院,浙江杭州310018)摘要:设备制造商为了满足设备用户对快速故障诊断和维护的需求,需要对设备进行实时监测、预测和远程指导。以数字孪生五维模型为基础,提出基于数字孪生的协同维护七维模型和维护服务协同模式。该模式下设备制造商和设备用户可实时获得对设备的监测信息,使得制造商能向用户提供更有效、及时的服务和支持。最后以某瓶装水封盖过程为例,构建设备隐式半马尔可夫模型,对设备当前运行状态的持续时间进行测算,得到与下一次劣化或故障状态的时间间隔,起到预测效果。构建了数字孪生服务平台框架和维护知识数据库,验证了所提出的协同维护模式的可行性。关键词:数字孪生;预测性维护;协同维护;隐式半马尔可夫模型;售后设备中图分类号:C931文献标识码:AResearchonCollaborativeModelofPredictiveMaintenanceforAfter-salesEquipmentBasedonDTLIXiao,CHENYuchen*,RUANYuanpeng,WANGLei,LAOXiaoyun(SchoolofManagement,HangzhouDianziUniversity,Hangzhou,Zhejiang310018,China)Abstract:Inresponsetothedemandsofusersforpromptfaultdiagnosisandmaintenance,equipmentmanufacturersrequiremoreadvancedmaintenancetechnologiesforreal-timemonitoring,prediction,andremoteguidance.Basedontheexistingfive-dimensionaldigitaltwinmodel,aseven-dimensionalmodelofcollaborativemaintenanceandacollaborativemaintenancemodelwereproposedbasedonthedigitaltwinvirtualrepresentation.Inthesemodels,equipmentmanufacturersanduserscouldobtainreal-timemonitoringinformationofequipment.Thisenabledmanufacturerstoprovidemoreeffectiveandtimelyserviceandsupporttocustomers.Finally,byusingabottledwatercappingprocessasanexampleandutilizingthehiddensemi-Markovmodelofequipment,thedurationofthecurrentoperationalstateforcappingequipmentwerepredicted.Thetimeintervalwiththenextdegradationorfaultstatewasobtained,whichhasthepredictioneffect.Adigitaltwinserviceplatformframeworkandamaintenanceknowledgedatabasewereconstructedtoverifythefeasibilityofthecollaborativemaintenancemodel.K...