基于特征提取的压力蒸汽灭菌设备运行故障诊断方法研究

Research on Fault Diagnosis Method of Pressure Steam Sterilization Equipment Based on Feature Extraction

目的针对现有压力蒸汽灭菌设备运行故障诊断中存在诊断准确度较低、耗时较长的问题,提出设计一种基于特征提取的压力蒸汽灭菌设备运行故障诊断方法。方法通过Lab VIEW编程软件结合数据采集卡,采集压力蒸汽灭菌设备运行故障数据。通过AR模型和敏感IMF分解方法提取压力蒸汽灭菌设备运行故障信号特征,并通过近邻图构建映射,利用拉普拉斯算子获取映射矩阵,将数据直接映射至低维子空间中,完成设备运行故障特征的约简。通过神经模糊推理自适应系统设计智能混合运行故障诊断模型,将约简后的特征数据输入模型中,实现压力蒸汽灭菌设备运行故障诊断。结果利用提出的设计方法对某医院压力蒸汽灭菌设备进行故障诊断,所设计方法对该设备故障的平均诊断准确率达到了89.2%,不同故障类型的平均诊断时间在640 ms内。结论本研究设计方法可有效提升压力蒸汽灭菌设备运行故障诊断效果,具有一定可行性。

Objective To solve the problems of low diagnostic accuracy and long time-consuming in the operation fault diagnosis of existing pressure steam sterilization equipment,a method of operation fault diagnosis of pressure steam sterilization equipment based on feature extraction is proposed.Methods The operation fault data of pressure steam sterilization equipment were collected through Lab VIEW programming software and data acquisition card.Through AR model and sensitive IMF decomposition method,the operation fault signal characteristics of pressure steam sterilization equipment were extracted,and the mapping was constructed through nearest neighbor graph.The mapping matrix was obtained by Laplace operator,and the data was directly mapped to low dimensional subspace to complete the reduction of equipment operation fault characteristics.The intelligent hybrid operation fault diagnosis model was designed through the neural fuzzy inference adaptive system,and the reduced characteristic data was input into the model to realize the operation fault diagnosis of pressure steam sterilization equipment.Results The design method was used to diagnose the fault of pressure steam sterilization equipment in a hospital.The average diagnosis accuracy of the designed method was 89.2%,and the average diagnosis time of different fault types was within 640 ms.Conclusion The designed method can effectively improve the operation fault diagnosis effect of pressure steam sterilization equipment,and has certain feasibility.