基于BP神经网络的螺杆泵井故障诊断技术研究

提出了基于BP神经网络和专家系统的螺杆泵井故障诊断方法:将螺杆泵井工况类型细分为10类,选取表征油井生产状态的7个特征参数作为输入量,采用Active X技术,借助VB调用Mat-lab人工神经网络工具箱函数,构建和训练网络模型.用该方法对新疆油田吉7井区吉101井和吉002井进行实例分析,验证了模型的正确性.

Fault diagnosis technique of screw pump wells based on BP neural network

In order to improve the economic benefits and management level of screw pump wells,a fault diagnosis technique of screw pump wells based on BP neural network and expert system is proposed.The working conditions of screw pump wells are classified into 10 types.Seven characteristic parameters in the production state of the oil wells are selected as the input variables of the BP neural network,and the BP neural network is established and trained by calling Matlab artificial neural network toolbox by VB through Active X technology.27 oil wells in Ji 7 wellblock of Xinjiang Oilfield are diagnosed using the trained BP neural network.The results show that: 7 oil wells appear pump leakage or stator swelling.Stator rubber is very sensitive to the oil of this wellblock,which leads to the stator rubber swelling and being damaged due to long time running,and then leads to pump leakage.This method provides important guidance for the field operations of the screw pump wells.