摘要
为了有效利用泵功图技术改善机械有杆采油系统采收率,对抽油泵不同工况所具有的不同泵功图进行图形特征分析,并利用改进算法的神经网络完成泵功图工况诊断.实例诊断结果与实际测试结果吻合,证明改进后的神经网络能够对不同工况的泵功图进行准确有效的特征聚类和模式识别.该方法具有工程应用价值.
In order to effectively utilize pump work indicating diagram for improving the efficiency of mechanical sucker-rod pumping system,the characteristics of the pump work indicating diagrams under different working conditions are analyzed,and the diagnosis of the working conditions of the mechanical sucker-rod pumping system is accomplished by means of the neural network of an improved BP algorithm.So the faults of the mechanical sucker-rod pumping system can be judged according to the characteristics of its pump work indicating diagrams.The diagnosing result of a case is identical to the measured result,which indicates the improved neural network can accurately identify and cluster the characteristics of the pump work indicating diagrams.
出处
《西安石油大学学报（自然科学版）》
CAS
2007年第3期119-121,共3页
Journal of Xi’an Shiyou University（Natural Science Edition）
关键词
BP神经网络
泵功图
灰度矩阵
故障诊断
模式识别
BP neural network
graph of pumpwork
gray-level matrix
fault diagnosis
pattern classification
