基于功率-位移图的调径变矩抽油机井工况诊断新模型

New model for diagnosing working conditions of adjustable-diameter and variable-torque pumping wells based on power-displacement diagram

基于自主研发调径变矩抽油机井8种典型工况的功率-位移图,利用灰色理论和曲线矩方法提取图形特征值,建立了典型功率-位移图特征值库;实测功率-位移图归一化处理后,提取特征值,并与特征值库进行灰色关联分析,根据最大关联度诊断油井工况。经现场90口油井实测功率-位移图工况诊断的检验,灰色理论诊断法诊断符合率为84.4%,曲线矩诊断法诊断符合率为88.9%,且部分工况类型诊断符合率较低。基于两种方法的诊断结果,对每种方法的单一工况诊断符合率进行分析,计算概率矩阵,建立了基于灰色理论和曲线矩的功率-位移图联合诊断模型,经90口油井工况诊断检验,总符合率为92.2%,且每种工况的诊断符合率均提升至90%以上,能够为调径变矩抽油机井工况智能诊断和生产优化提供技术支持。

Based on the independently developed power-displacement diagrams of adjustable-diameter variable-torque pumping wells in 8 typical working conditions,the characteristic values of the graphs were extracted through grey theory and curve moment method,to establish the corresponding characteristic value library of the typical power-displacement diagram.After the measured power-displacement diagram was normalized,the characteristic value was extracted,and the grey correlation analysis was performed with the characteristic value library to diagnose the oil well working condition according to the maximum correlation degree.According to the diagnostic test of 90 oil wells,the diagnostic coincidence rate of the grey theory diagnosis method was 84.4%,the diagnosis coincidence rate of the curve moment diagnosis method was 88.9%,and the diagnosis coincidence rate of some working condition types was relatively low.Based on the diagnosis results of the two methods,it was analyzed the coincidence rate of the single working condition diagnosis of each method,the probability matrix was calculated,and it was established the power-displacement diagram joint diagnosis model based on grey theory and curve moments.After the tests of 90 oil wells,the total coincidence rate was 92.2%,and the diagnosis coincidence rate of each working condition was improved to more than 90%,which can provide technical support for intelligent diagnosis and production optimization of adjustable-diameter and variable-torque pumping wells.