频段能量统计法在机械设备工作状态监测中的应用  

Application of Frequency Band Energy Statistics Method to Monitoring the Work Conditions of Mechanical Equipment

作　　者：袁云龙[1] 袁超勇[2] 杨超珍[1] 

机构地区：[1]宁波工程学院机械学院,浙江宁波315016 [2]爱立迈科(宁波)计测仪器有限公司,浙江宁波315040

出　　处：《轻工机械》2010年第2期85-88,共4页Light Industry Machinery

基　　金：国家高技术研究发展计划(863计划;2008AA100905)

摘　　要：提出了一种对振动信号频谱新的分析方法,该方法通过对机械设备振动信号进行分段多次FFT变换和1/3倍频程谱计算,并对特征频率段能量值作直方图图谱,获得其能量分布特征,与处于正常工作状态的对应能量直方图图谱作偏移比较,从而实现对机械设备工作状态监测与故障诊断。结果表明,该方法对振动信号的频谱变化具有相当强的敏感性,能有效地分辨机械设备工作状态的一些微小变化,为机械设备的工作状态监测或故障诊断提供了1条新的准确、高效的途径。A new analysis method frequency band energy statistics method, has been proposed for the analysis of the spectra from the vibration signals. The recorded vibration signals were first divided into multi-segment, then each segment was calculated via the FFT transformation and 1/3 octave spectrum to make out histogram drawings with features of the frequency energy value and obtained the characteristies of energy distribution. Consequently, we can monitor the work conditions and fault diagnosis for the mechanical equipment by the deviation compared analysis of the corresponding histogram drawings of equipment with normal work conditions. The results show that present method exhibits a very strong sensitivity to the changes of vibration signal, leading to the fine detection of the minor changes from the equipment work conditions, thus might provide a novel precise and high efficient method for definitely monitoring the work conditions and fault diagnosis of the mechanical equipment.

关 键 词：机械振动 故障诊断 频谱分析 能量偏移 

分 类 号：TB123[理学—工程力学]