融合D-S证据理论与BP算法的机械设备故障诊断系统研发  

Development of Mechanical Equipment Fault Diagnosis System Integrated with D-S Evidence Theory and BP Algorithm

作　　者：段浩 张方琪 DUAN Hao;ZHANG Fangqi(No.987 Hospital of Joint Logistic Support Force,Baoji Shaanxi 721004,China)

机构地区：[1]联勤保障部队第九八七医院,陕西宝鸡721004

出　　处：《自动化与仪器仪表》2025年第3期172-176,共5页Automation & Instrumentation

基　　金：市级,联勤保障部队987医院一般技术创新项目(2023DJBQYY-16)。

摘　　要：医疗机械设备在疾病诊断和治疗中占有重要地位。然而,传统人工诊断难以满足医疗机械设备高精密度和复杂性带来的高诊断要求。因此,研究提出一种融合邓普斯特-谢弗证据理论与反向传播算法的故障诊断策略,并将其应用于研发医疗机械设备故障诊断系统中。通过实验,研究确定了反向传播算法的测试函数与学习率。进一步的实验结果表明,研究研发的医疗机械设备故障诊断系统在诊断精确率、稳定性和实时性方面均优于现有系统,其平均精确率达到98.46%,平均误报率低至1.01%,平均诊断时间仅为120.2 ms,显示出优越的性能。研究不仅能够提高诊断准确性和效率,还能够通过提出的机械设备故障诊断系统在未来有效减少故障对医疗服务的影响,保障患者安全。Medical machinery and equipment play an important role in disease diagnosis and treatment.However,traditional manual diagnosis can hardly meet the high diagnostic requirements brought by the high precision and complexity of medical machinery and equipment.Therefore,this study proposes a fault diagnosis strategy that integrates Dempster-Schafer evidence theory and back propagation algorithm,and applies it to the research and development of medical machinery and equipment fault diagnosis system.Through experiments,the study determined the test function and learning rate of the back propagation algorithm.Further experimental results show that the medical machinery and equipment fault diagnosis system developed in this study is superior to the existing system in terms of diagnostic accuracy,stability and real-time performance.Its average accuracy rate reaches 98.46%,the average false alarm rate is as low as 1.01%,and the average diagnosis time is only 120.2 ms,showing superior performance.The research can not only improve the diagnostic accuracy and efficiency,but also effectively reduce the impact of faults on medical services and ensure patient safety through the proposed mechanical equipment fault diagnosis system in the future.

关 键 词：医疗 机械设备 故障诊断 D-S证据理论 BP算法 

分 类 号：TH16[机械工程—机械制造及自动化]