A Fusion Method Based on Deep Learning for Fault Diagnosis of Oilfield Injection Pumps with Unbalanced Data

doi:10.16183/j.cnki.jsjtu.2024.308

Abstract

Abstract:

To address the data imbalance problem in the fault diagnosis of oilfield plunger injection pumps, a multi-level Inception-long short-term memory (LSTM) network model integrated with wavelet packet decomposition (WPD) and efficient channel attention (ECA) mechanism is proposed. The model utilizes WPD technology to decompose the low-frequency and high-frequency components of vibration signals. The Inception module extracts multiple-scale data features, while the LSTM module captures temporal correlations in the data. Furthermore, the ECA mechanism further enhances the model’s capability to exploit cross-channel data correlations, thereby improving the accuracy of feature representation. Experiments are conducted using plunger pump vibration data collected from an actual oilfield operation site. The results show that the proposed model achieves optimal performance with a diagnostic accuracy of 99.38%, which demonstrates its effectiveness and superiority.

Key words: plunger injection pump, imbalance data, convolutional neural network (CNN), long short-term memory (LSTM) network, attention mechanism

CLC Number:

TP181

WU Zelin, LUO Feng, CUI Xiwen, CHENG Xin, XIA Tangbin. A Fusion Method Based on Deep Learning for Fault Diagnosis of Oilfield Injection Pumps with Unbalanced Data[J]. Journal of Shanghai Jiao Tong University, 2026, 60(6): 1017-1025.

Figures/Tables 15

Fig.1

Fig.2

Fig.3

Fig.4

Tab.1

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

Tab.2

Fig.10

Fig.11

Fig.12

Tab.3

References 20

[1]	HE X, XIAO G, HU B, et al. The applications of energy regeneration and conversion technologies based on hydraulic transmission systems: A review[J]. Energy Conversion and Management, 2020, 205: 112413. doi: 10.1016/j.enconman.2019.112413 URL
[2]	SHAO K, FU W, TAN J, et al. Coordinated approach fusing time-shift multiscale dispersion entropy and vibrational Harris Hawks optimization-based SVM for fault diagnosis of rolling bearing[J]. Mea-surement, 2021, 173: 108580.
[3]	XIA T, JIANG Y, DING Y, et al. Intelligent maintenance framework for reconfigurable manufacturing with deep learning-based prognostics[J]. IEEE Internet of Things Journal, 2024, 11(13):22853-22868. doi: 10.1109/JIOT.2024.3357750 URL
[4]	JIANG Y, XIA T, WANG D, et al. Joint adaptive transfer learning network for cross-domain fault diagnosis based on multi-layer feature fusion[J]. Neurocomputing, 2022, 487: 228-242. doi: 10.1016/j.neucom.2021.11.005 URL
[5]	JIANG G, HE H, YAN J, et al. Multiscale convolutional neural networks for fault diagnosis of wind turbine gearbox[J]. IEEE Transactions on Industrial Electronics, 2018, 66(4):3196-3207. doi: 10.1109/TIE.2018.2844805 URL
[6]	DU J, WANG S, ZHANG H. Layered clustering multi-fault diagnosis for hydraulic piston pump[J]. Mechanical Systems and Signal Processing, 2013, 36(2): 487-504. doi: 10.1016/j.ymssp.2012.10.020 URL
[7]	LAN Y, HU J, HUANG J, et al. Fault diagnosis on slipper abrasion of axial piston pump based on extreme learning machine[J]. Measurement, 2018, 124: 378-385. doi: 10.1016/j.measurement.2018.03.050 URL
[8]	刘秀丽, 徐小力. 基于特征金字塔卷积循环神经网络的故障诊断方法[J]. 上海交通大学学报, 2022, 56(2): 182-190. doi: 10.16183/j.cnki.jsjtu.2021.001
	LIU Xiuli, XU Xiaoli. A fault diagnosis method based on feature pyramid CRNN network[J]. Journal of Shanghai Jiao Tong University, 2022, 56(2): 182-190.
[9]	卓鹏程, 严瑾, 郑美妹, 等. 面向滚动轴承全生命周期故障诊断的GA-OIHF Elman神经网络算法[J]. 上海交通大学学报, 2021, 55(10): 1255-1262. doi: 10.16183/j.cnki.jsjtu.2020.157
	ZHUO Pengcheng, YAN Jin, ZHENG Meimei, et al. GA-OIHF Elman neural network algorithm for fault diagnosis of full life cycle of rolling bearing[J]. Journal of Shanghai Jiao Tong University, 2021, 55(10): 1255-1262.
[10]	马航宇, 周笛, 卫宇杰, 等. 变工况下基于自适应深度置信网络的轴承智能故障诊断[J]. 上海交通大学学报, 2022, 56(10): 1368-1377. doi: 10.16183/j.cnki.jsjtu.2021.161
	MA Hangyu, ZHOU Di, WEI Yujie, et al. Intelligent bearing fault diagnosis based on adaptive deep belief network under variable working conditions[J]. Journal of Shanghai Jiao Tong University, 2022, 56(10): 1368-1377.
[11]	QIAN J, SONG Z, YAO Y, et al. A review on autoencoder based representation learning for fault detection and diagnosis in industrial processes[J]. Chemometrics and Intelligent Laboratory Systems, 2022, 231: 104711. doi: 10.1016/j.chemolab.2022.104711 URL
[12]	YANG J, LIU J, XIE J, et al. Conditional GAN and 2-D CNN for bearing fault diagnosis with small samples[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-12.
[13]	FAN S, ZHANG X, SONG Z. Imbalanced sample selection with deep reinforcement learning for fault diagnosis[J]. IEEE Transactions on Industrial Informatics, 2021, 18(4): 2518-2527. doi: 10.1109/TII.2021.3100284 URL
[14]	LIU S, CHEN J, QU C, et al. LOSGAN: Latent optimized stable GAN for intelligent fault diagnosis with limited data in rotating machinery[J]. Measurement Science and Technology, 2021, 32(4): 045101. doi: 10.1088/1361-6501/abd0c1
[15]	ZHAO B, YUAN Q. Improved generative adversarial network for vibration-based fault diagnosis with imbalanced data[J]. Measurement, 2021, 169: 108522. doi: 10.1016/j.measurement.2020.108522 URL
[16]	SZEGEDY C, LIU W, JIA Y, et al. Going deeper with convolutions[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston, USA: IEEE, 2015: 1-9.
[17]	WANG Q, WU B, ZHU P, et al. ECA-Net: Efficient channel attention for deep convolutional neural networks[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, USA: IEEE, 2020: 11534-11542.
[18]	VAN HOUDT G, MOSQUERA C, NÁPOLES G. A review on the long short-term memory model[J]. Artificial Intelligence Review, 2020, 53(8): 5929-5955. doi: 10.1007/s10462-020-09838-1
[19]	PANDEY S K, JANGHEL R R. Automatic detection of arrhythmia from imbalanced ECG database using CNN model with SMOTE[J]. Australasian Physical and Engineering Sciences in Medicine, 2019, 42(4): 1129-1139. doi: 10.1007/s13246-019-00815-9 pmid: 31728941
[20]	CUI Y, JIA M, LIN T, et al. Class-balanced loss based on effective number of samples[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019: 9268-9277.

传感器序号	传感器位置	传感器序号	传感器位置
1	底座东南	9	电机东
2	底座东北	10	电机西
3	底座西北	11	曲轴东
4	底座西南	12	泵头正上方
5	曲轴轴承	13	泵头正面
6	西柱塞密封函	14	泵进口管线
7	中柱塞密封函	15	泵出口管线
8	东柱塞密封函

卷积层	k	卷积层	k
1	5	3-1	5
2-1	5	3-2	7
2-2	7	3-3	9
2-3	9

模型	准确率/%	精确率/%	召回率/%
CNN	93.33	90.26	93.50
CB-CNN	96.14	95.83	96.68
SMOTE-CNN	94.10	92.37	94.19
CNN-LSTM	95.79	94.01	95.49
Inception-LSTM	96.31	95.14	97.17
Inception-LSTM-ECA	98.82	98.49	98.81
CNN-LSTM-WE	98.39	96.94	99.03
Inception-WE	97.84	98.42	96.35
Inception-LSTM-WPD	97.77	97.25	98.42
MILN-WE	99.38	99.42	99.27