抽油机井智能诊断系统

NEW INTELLIGENT TECHNOLOGIES FOR DIAGNOSTIC THE FAULT OF SUCKER ROD PUMPING SYSTEM

赵凤芝 包锋 刘贤梅

Zhao Fengzhi Bao Feng Liu Xianmei

大庆石油学院计算机科学系, 黑龙江安达 151400

为快速、准确地判断抽油井井下工作状况,将人工神经网络、常规数学计算及传统知识推理等多种诊断方法集成一体,研制出一套集成化的抽油机井工况智能故障诊断系统,并论述了该项技术实现的技术思路、系统结构、功能、具体实现原理。与传统的技术相比具有识别速度快、准确率高等特点,为实现抽油机井智能化故障诊断提供了一种新途径。

In order to minimize operating costs and maximize oil production, a quick and accurate identification of down-hole problem is essential in rod pumping system. This paper proposed a new integrated intelligent technology for diagnostic the fault of the sucker rod pumping system, which is integrated conventional knowledge inference, artificial neural networks and conventional mathematics algorithm, and in detail, introduced the whole developing procedure of an intelligent diagnosis system, including system designing, composing, function and realization. This method can do a faster and better job than conventional technologies. It is presented a new way for intelligent diagnostic the fault of the sucker rod pumping system.

10.3969/j.issn.1000-7393.2000.03.021