利用时域参数联合统计钻井泵振动特征提取的研究
Research on Vibration Feature Extraction of Drilling Pump based on Time Domain Joint Statistical Parameters

开展机械设备振动信号时域统计参数信息挖掘的研究，为检测、诊断设备故障，预测设备的工作状态服务，是一项很有意义的工作。以钻井泵为研究对象，对振动信号进行基元分段处理，统计基元分段后振动信号的时域参数；采用一种新的框架-时域参数联合统计分布，在时域参数联合坐标系下详细考察振动信号时域统计参数的分布特征与量化规律。研究表明，在不同工况下时域参数散点的分布具有很强的规律性与空间区域特征，能够在时域波形信号、参数分布特征和分布区域，以及设备故障之间建立很好的对应关系。
Research on mining combined time domain statistical parameter information from vibration signal is a very meaningful work for machinery detection, fault diagnosis and state prediction service. In this paper, drilling pump's vibration signal is segmented by BOU(Basic operation unit) processing, then time domain parameters of such vibration signal's segmentation has been calculated. These time domain parameters are investigated carefully with using a new framework-statistical distribution of combined time domain parameters. Studies have shown that the distribution of time domain parameter scatters has strong regularity and spatial feature under different working conditions. A good corresponding relationship between the vibration signal, combined parameter distribution characteristics, regional distribution, and equipment malfunction has been established. It can also provide a geat convenience for the fault detection of drilling pump