摘要
为了提高离心泵故障诊断的准确性,简化诊断的操作流程,本文提出了一种共振特征识别法。首先,依据离心泵的共振特征,深入分析离心泵状态、共振异常值,形成特征集合;其次,利用历史数据、共振情况、过热预警等函数不断调整共振特征,形成共振识别图谱;最后,结合泵机的稳定性指标,验证故障诊断结果。仿真结果显示,本文提出的共振特征识别法能提高离心泵的共振特征识别率,缩短故障诊断时间,显著优于人工检查法。共振特征识别法能满足离心泵的故障诊断要求。
In order to improve the accuracy of centrifugal pump fault diagnosis and simplify the diagnosis process,this paper proposes a resonance feature recognition method.First o£all,according to the resonance characteristics of centrifugal pump,the state of centrifugal pump and the abnormal value of resonance are analyzed in depth to form a feature set.Then,the resonance characteristics are continuously adjusted with historical data,resonance conditions,overheat warning and other functions to form a resonance recognition map.Finally,combined with the stability index of the pump,the fault diagnosis results are verified.The simulation results show that the resonance feature recognition method proposed in this paper can improve the resonance feature recognition rate of centrifugal pump and shorten the fault diagnosis time,which is significantly better than the manual inspection method.The resonance feature recognition method can meet the fault diagnosis requirements of centrifugal pump.
作者
王治国
吴茜
WANG Zhiguo;WU Qian(Sinopec Zhongyuan Petroleum Engineering Co.,Ltd.Tarim Branch,Korla,Xinjiang 841000,China;Oil and Gas Storage and Transportation Center of Sinopec Zhongyuan Oilfield Branch,Puyang,Henan 457001,China)
出处
《自动化应用》
2023年第7期144-146,共3页
Automation Application
关键词
共振
特征
算法
诊断
离心泵
仿真
resonance
features
algorithm
diagnosis
centrifugal pump
simulation