基于数字孪生技术的医用制氧机预测性维护方法研究OACSTPCD

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.