基于电性能特征的医疗设备故障识别方法研究

Research on the method of fault identification based on the features of electrical performance of medical equipment

目的:提出一种多电性能特征提取的医疗设备故障识别方法,研究医疗设备输出的电流、电压和功率3种电信号,从而建立一套故障知识库,是一种用于替代人工检测的智能化医疗设备检测方法。方法:采集一种医疗设备的输出电信号进行分析,搭建电信号在正常与故障状态下的数学模型,并提取与状态相关的特征,通过建立设备工作状态、故障事件与电信号特征之间的关系而得到故障识别方法。设计两种实验验证了该方法的有效性和准确性:①采用不同类别医疗设备的特征研究,证明了该方法具备电子医疗设备的普适性;②采集同类医疗设备的不同状态和故障事件,验证不同设备的信号特征提取与理论模型的一致性,同时提取同种设备的故障状态下的电性能特征,用于制定相应故障规则后与人工故障检测准确性和时间进行对比。结果:采用多电性能特征提取的医疗设备故障识别方法,在对超声诊断仪和监护仪设备样本检测的验证中,仅电流特征的故障识别率普遍>90%;并且计算机通过制定的规则识别故障所需时间与人工检测的时间相比节省70%。结论:多电性能特征提取的医疗设备故障识别方法可用于医疗设备故障的识别,为有效提升医疗设备故障检测效率奠定基础。

Objective:To propose a method of fault identification based on the extraction of multi electrical performance feature of medical equipment through researched the outputted 3 types of electrical signals from medical equipment,so as to establish a set of fault knowledge base,which was one new detection method that could replace the manual detection of intelligent medical equipment.Methods:The outputted electrical signal of one kind of medical equipment was collected and analyzed in this research.The mathematical model of electrical signals in normal and fault state was constructed,and the features with the relative state were extracted.The method of fault identification was obtained through established the relationship between working state of equipment,fault event and electrical signal features.After that,two experiments were designed to verify the effectiveness and accuracy of the method.①The first experiment adopted the feature studies of different medical equipment to prove the method opposed the universality of electronic medical equipment.②The second experiment collected different states and fault events of similar medical equipment to verify the consistency between signal feature extraction from different equipment and the theoretical model.At the same time,the electrical performance features of the same kind of equipment under the fault state were extracted and were used to compare the accuracy and the time between manual detection and the new detection method after the corresponding fault rule was formulated.Results:Only the fault identification rate of the electric feature was generally larger than 90%after the method of identifying fault of medical equipment,which adopted multifeature of electrical performance to extract features,was adopted in the verifications for the detections of ultrasonic diagnosis device and monitoring device.In addition,the spent time of computer through the formulated rules to identify fault saved 70%of the time of manual detection.Conclusion:The method with the multi electric performance features of identifying fault of medical equipment can be used in the identification for the fault of medical equipment,which lays foundation for the effective improvement of the detection efficiency of the fault of medical equipment.