基于可靠性分析与状态监测的钻井泵剩余工作寿命预测  

Drilling pump remaining working life prediction based on reliability analysis and condition monitoring

作　　者：裴峻峰[1] 张嗣伟[2] 齐明侠[3] 张志毅[4] 

机构地区：[1]常州大学机械与能源工程学院,江苏常州213016 [2]中国石油大学机械与储运工程学院,北京102249 [3]中国石油大学机电工程学院,山东东营257061 [4]河南石油勘探局钻井工程公司,河南南阳473132

出　　处：《中国石油大学学报（自然科学版）》2010年第5期121-125,共5页Journal of China University of Petroleum(Edition of Natural Science)

基　　金：江苏省科技成果转化专项资金项目(BA2005041);江苏省油气储运技术重点实验室建设项目(CY0606)

摘　　要：在对大量同类钻井泵的大修间隔时间数据进行概率统计分析的基础上,求得与钻井泵的运行可靠度相关的特征参数,在宏观上提供钻井泵寿命预测的依据。动力端中情况最差的轴承的寿命可代表钻井泵的寿命,所以将状态监测所得的经优选的各域振动信息作为神经网络系统的模糊输入向量,通过径向基神经网络求出轴承的故障隶属度,作为轴承理论寿命计算公式的修正系数,由此解决钻井泵剩余工作寿命的预测问题,进而获得钻井泵剩余工作寿命。实际预测结果证明了预测方法的科学性和合理性。Based on probability and statistics analysis of a large number of overhaul interval time data of similar drilling pumps,the characteristic parameters related to the operating reliability of the drilling pump were obtained. These parameters provide macroscopic basis for life prediction of drilling pump. Because the life of the worst bearing at the power end represents the life of the drilling pump,each optimal domain vibration information obtained from condition monitoring was used as fuzzy input vectors of a neural network system,then the value of fault membership degree of the bearing was calculated through the RBF neural network. The value acts as the correction factor of theoretical bearing life calculation formula. As a result,the remaining working life of the drilling pump can be determined. The actual predicted results show the scientific and reasonable prediction method.

关 键 词：钻井泵 动力端轴承 可靠性分析 状态监测 剩余工作寿命预测 

分 类 号：TE926[石油与天然气工程—石油机械设备] N945.24[自然科学总论—系统科学]