基于信号分析技术的小儿呼吸机工作状态自动化识别研究  

Research on Automatic Recognition of Working Status of Pediatric Respirators Based on Signal Analysis Technology

作　　者：李园园 LI Yuanyuan(Shengjing Hospital of China Medical University,Shenyang 110000,China)

机构地区：[1]中国医科大学附属盛京医院,沈阳110000

出　　处：《自动化与仪器仪表》2024年第12期299-303,共5页Automation & Instrumentation

摘　　要：针对当前小儿呼吸机在工作状态下的故障诊断难题,研究提出了一种信号分析技术基础上的自动化识别方法。通过利用Prony分析进行故障特征信号的提取,并利用支持向量机进行信号识别与分类。Prony分析显示,当小儿呼吸机在工作状态下发生故障时,其信号幅频谱图的直流分量幅度显著增加。性能验证显示,研究提出的识别方法曲线下面积值比其他方法平均增加了18.67%。故障信号的识别时间比反向神经网络减少了35.71%。在识别准确性方面,研究提出的识别方法在1 000个样本信号的故障识别中准确性高达96.15%。结果表明,研究提出的小儿呼吸机工作状态自动化识别方法对故障特征识别准确性和识别效果更优越,在小儿呼吸机等医疗设备工作过程中的故障监控具有一定的应用价值。A signal analysis technology based automated identification method is proposed to address the current problem of diagnosing faults in pediatric ventilators during operation.By using Prony analysis to extract fault feature signals,and using support vector machines for signal recognition and classification.Prony analysis shows that when a pediatric ventilator malfunctions during operation,the amplitude of the DC component in its signal amplitude spectrum significantly increases.Performance verification shows that the proposed recognition method has an average increase of 18.67%in area under the curve compared to other methods.The recognition time of fault signals is reduced by 35.71%compared to reverse neural networks.In terms of recognition accuracy,the proposed recognition method achieved an accuracy of 96.15%in fault recognition of 1000 sample signals.The results indicate that the proposed automatic recognition method for the working status of pediatric ventilators has superior accuracy and recognition effect in fault feature recognition,and has certain application value in fault monitoring during the working process of medical equipment such as pediatric ventilators.

关 键 词：小儿呼吸机 信号分析技术 PRONY分析 小波变换 状态识别 

分 类 号：TN98[电子电信—信息与通信工程] TP39[自动化与计算机技术—计算机应用技术]