针对电子设备间歇性故障诊断的难题,提出一种基于声音传感器的电子设备间歇性故障诊断方法(Intermittent Fault Diagnosis Method for Electronic Devices Based on Acoustic Sensors,IFDM-AS).IFDM-AS利用微机电系统(Micro-Electro-Mechanical System,MEMS)麦克风采集设备运行声音,通过梅尔频率倒谱系数(Mel-Frequency Cepstral Coefficients,MFCC)特征提取、支持向量机(Support Vector Machine,SVM)模式识别、贝叶斯网络故障诊断决策技术,实现设备健康状态的实时诊断.实验结果表明,IFDM-AS在5种电子设备状态下的平均识别准确率为96.7％,有效性和可靠性较强.
Intermittent Fault Diagnosis Method for Electronic Devices Based on Acoustic Sensors
For the problem of intermittent fault diagnosis of electronic equipment,an Intermittent Fault Diagnosis Method for Electronic Devices Based on Acoustic Sensors (IFDM-AS) is proposed. IFDM-AS uses Micro-Electro-Mechanical System (MEMS) microphone acquisition equipment to run sound through Mel-Frequency Cepstral Coefficients (MFCC) feature extraction,Support Vector Machine (SVM) pattern recognition,bayesian network fault diagnosis and decision technology,to achieve real-time diagnosis of equipment health status. The experimental results show that the average recognition accuracy of IFDM-AS is 96.7％ under five electronic device states,and the effectiveness and reliability are strong.
万方数据
维普
