基于神经网络的示功图诊断研究  

Diagnostic of Indicator Diagram Research Based on Neural Network

作　　者：肖维民[1] 胡锦[1] 

机构地区：[1]安徽工业大学计算机系,安徽马鞍山243000

出　　处：《工业控制计算机》2012年第8期68-69,共2页Industrial Control Computer

摘　　要：在抽油机井生产过程中,及时掌握抽油机井工况,实现抽油机井的远程监测和科学管理,是当前抽油机采油系统迫切需要研究的一个重大课题。在传统工况诊断的基础上研究了抽油机井工况智能诊断技术。通过对示功图的网格化,建立了示功图的计算模型,通过对这些模型进行预处理,建立了抽油机井典型工况的特征参数样本库。研究了BP神经网络的基本结构,并将BP神经网络诊断系统应用于抽油机井示功图识别,取得了良好的效果。In the production process of the pumping wells,to grasp the working conditions of the pumping wells,to achieve remote monitoring of pumping wells and scientific management is a major topic which needs to be studied urgently in current pumping system.This paper studies intelligent diagnostic technique of pumping wells based on traditional diagnostic of working condition.First,this paper make the Indicator diagram gridded,then by pretreatment of these models,establish a sample library of the characteristic parameters of typical operating conditions of pumping wells.Finally,the basic structure of BP neural network and BP neural network diagnostic system is applied to the pumping wells to identify the indicator diagram.

关 键 词：BP神经网络 示功图 网格化 

分 类 号：TE933.1[石油与天然气工程—石油机械设备]