基于数字孪生技术的医用制氧机预测性维护方法研究  

Study on Predictive Maintenance for Medical Oxygen Generator Based on Digital Twin Technology

作　　者：刘济浔 蒋益钢 左建雄 裘超 张星 LIU Jixun;JIANG Yigang;ZUO Jianxiong;QIU Chao;ZHANG Xing(School of Management,Zhejiang Shuren University,Hangzhou Zhejiang 310015,China;Research&Development Center,Hangzhou Medoxygen Technology Co.,Ltd.,Hangzhou Zhejiang 311122,China;Department of Medical Engineering,Zhejiang Qiushi Cardiovascular Hospital,Hangzhou Zhejiang 310012,China)

机构地区：[1]浙江树人学院管理学院,浙江杭州310015 [2]杭州颐氧健康科技有限公司技术研究中心,浙江杭州311122 [3]浙江求是心血管病医院医学工程部,浙江杭州310012

出　　处：《中国医疗设备》2024年第12期46-52,共7页China Medical Devices

基　　金：国家卫生健康委医院管理研究所“2022公立医院后勤精细化管理”研究项目(GYZ2022HQ45)。

摘　　要：目的探讨一种基于数字孪生的方法,以提高医用制氧机的全寿命周期性能,并实现预测性维护。方法将数字孪生技术贯穿医用制氧机的工艺和系统设计以及运行的全过程,利用“数据+模型+算法”破解医用制氧机在智能运维过程中可能遭遇的状态变化、风险失控和成本上扬等难题。采用6步法构建医用制氧机虚拟的数字孪生体,应用多源信息融合监测评估技术、设备在线自适应精确诊断技术、设备运维风险预测与评估技术等,完善医用制氧机的数据采集和运行范式;同时通过安装在医用制氧机上的传感器收集的数据和部署在“颐氧云”服务器上的机器学习模型和算法,实现基于数字孪生技术的预测性维护。结果共有43个项目合计105台医用制氧机采用数字孪生体进行预测性维护,总计运行时间为214.8万小时,非计划停机率从部署前0.283%下降到0.029%,而单位时间备件成本从0.55元/h降至0.31元/h。经统计分析发现,数字孪生体前后非计划停机率和单位时间备件成本差异均有统计学意义(P<0.05)。结论基于数字孪生技术的预测性维护方法,可降低非计划停机率,从而提高医用制氧机全寿命周期性内的可靠性,进而延长设备的使用寿命,降低设备的使用成本。Objective To explore a method based on digital twin,to improve the life cycle performance of medical oxygen generator and achieve its predictive maintenance.Methods The digital twin technology was used throughout the whole process of the technogy and system design and operation of the medical oxygen generator,and the“data+model+algorithm”was used to address the problems such as condition changes,risk loss and cost escalation that the medical oxygen concentrator would encounter in the process of intelligent operation and maintenance.The 6-step method was adopted to create a virtual digital twin of medical oxygen generator,and the multi-source information fusion,equipment online accurate adaptive diagnosis,prediction and evaluation of operation and maintenance risks were applied to improve the data acquisition and operation paradigm of medical oxygen generator.Meanwhile,through the data collected by the sensors installed on the medical oxygen generator and the machine learning model and algorithm deployed on the“Medox”server,the predictive maintenance based on the digital twin technology was realized.Results Digital twin based predictive maintenance was deployed for total 43 projects with a total of 105 sets medical oxygen generators.The total operating time was 2.148 million hours,and the unplanned downtime rate had decreased from 0.283%to 0.029%and the spare part cost per unit time from 0.55 yuan/h down to 0.31 yuan/h.The statistical analysis showed that there was a significant difference in unplanned downtime rate and spare part cost per unit time after implementing digital twin based predictive maintenance(P<0.05).Conclusion Digital twin based predictive maintenance reduces unplanned downtime and improves the reliability of medical oxygen generators in its lifespan.And it can extend the service life of the equipment and reduce its operating costs.

关 键 词：医用制氧机 数字孪生 预测性维护 全寿命周期性能 可靠性分析 智慧运维 管理效能 

分 类 号：R197.39[医药卫生—卫生事业管理]