基于小波包能量和CS-ELM的有杆泵抽油井故障诊断  

Fault Diagnosis of Sucker Rod Pumping Wells Based on Wavelet Packet Energy and CS-ELM

作　　者：严永进 袁春华 

机构地区：[1]沈阳理工大学自动化与电气工程学院,辽宁沈阳110159

出　　处：《工业控制计算机》2024年第7期111-113,共3页Industrial Control Computer

摘　　要：针对目前有杆泵抽油井故障诊断人工分析方法效率低问题,提出一种基于小波包能量和布谷鸟算法优化极限学习机(CS-ELM)的有杆泵抽油井故障诊断方法。首先对电功率数据进行小波包分解得到多个子频带,计算各频带的能量值再归一化处理后组合成特征向量;然后通过CS算法优化ELM使其得到最优的输入权值和隐含层阈值;最后利用CSELM模型对提取的特征向量进行有杆泵抽油井故障诊断并与SVM、BP和ELM的诊断结果进行对比。仿真结果表明:CSELM的有杆泵抽油井故障诊断的准确率最高,验证了该方法的有效性。Currently,the manual analysis method for fault diagnosis of rod-pumped oil wells is inefficient,a fault diagnosis method of sucker rod pumping wells based on wavelet packet energy and cuckoo algorithm optimized extreme learning machine(CS-ELM)is proposed in this paper.Firstly,the electric power data is decomposed by wavelet packet to obtain multiple sub-bands,and the energy values of each band are calculated and normalized to form a feature vector.Then,the CS algorithm is used to optimize the ELM to obtain the optimal input weight and hidden layer threshold.Finally,the CS-ELM model is used to diagnose the fault of the rod pumping well with the extracted feature vector and compared with the diagnosis results of SVM,BP and ELM.

关 键 词：小波包能量 故障诊断 极限学习机 布谷鸟算法 

分 类 号：TE933.3[石油与天然气工程—石油机械设备] TP18[自动化与计算机技术—控制理论与控制工程]