基于泵送频率的往复泵活塞故障诊断方法

Fault diagnosis method based on pump stroke frequency for reciprocating pump piston

为了确保往复式钻井泵的高质量运行,实施在线故障监测至关重要。从BW-250型钻井泵液力端的易损件入手,设计了试验探究钻井泵的振动信号的成分;通过对频率成分的分析,揭示了泵送频率幅值与活塞刺漏故障之间的关系,并依据机理提出了以泵送频率幅值作为诊断指标的钻井泵活塞故障检测方法,进一步结合最大相关峭度解卷积滤波及包络谱分析等方法从泵送频率处能量变化的角度,对活塞刺漏这一故障进行了诊断;结合实验室以及钻井现场采集的数据对该方法进行了验证,并将诊断结果同其他振动指标做了对比。结果表明,该方法对于往复泵活塞故障诊断的准确率为91.1%,相较于均方根诊断准确率提升了3.6%、相较于脉冲因子、裕度因子和峭度3种统计指标诊断准确率提升了9%以上。该方法取得了良好的结果,为往复泵活塞组件的故障诊断提出了一种较好的解决思路。

Online fault monitoring is critical to ensure the high-quality operation of reciprocating drilling pumps.Firstly,an experiment was designed to explore the components of the vibration signals of the drilling pump by examining the easily worn parts of the hydraulic end of the BW-250 drilling pump.Then,the correlation between the amplitude of the pump delivery frequency and piston leakage faults was revealed by analyzing frequency components.Based on this mechanism,a method for detecting piston faults in drilling pumps using the pump delivery frequency amplitude as a diagnostic indicator was proposed.Furthermore,by integrating Maximum Correlated Kurtosis Deconvolution filtering and envelope spectrum analysis methods,the piston leakage fault was diagnosed from the perspective of energy changes at the pump delivery frequency.Finally,the method was validated using the data collected at laboratory and drilling field,and the diagnostic results were compared with other vibration indicators.The results show that the accuracy of this method for diagnosing piston faults in reciprocating pumps is 91.1%,an increase of 3.6% compared to the RMS index diagnostic accuracy and an increase of over 9% compared to impulse,margin and kurtosis index diagnostic accuracy.The method achieved good results,which provides a solution for fault diagnosis of reciprocating mechanical piston components.