采油过程多尺度状态特征生成的有杆泵动态液面预测

doi:10.11949/0438-1157.20190352

摘要/Abstract

摘要：

动态液面是反映油藏供液能力的重要参数，生产现场多采用回声仪进行检测，其检测效率低，无法连续检测，存在安全隐患。针对现有基于数据驱动的预测方法由于各种原因造成历史数据不足，进而导致建模困难的问题，通过对机理进行分析，找出与动态液面关系密切的多尺度状态特征；采用生成对抗网络生成状态特征，解决历史数据不足的问题；仿真实验表明生成的数据可以用于建立动态液面预测模型，并确定采用支持向量回归作为建模方法；最后在油田生产现场使用本文提出的建模方法，实际预测结果表明本方法的有效性，满足油田工程应用要求。

关键词: 动态液面, 软测量, 多尺度状态特征, 生成对抗网络, 支持向量回归

Abstract:

Dynamic liquid level is an important parameter to reflect the liquid supply capacity of oil deposit. In the actual production site, dynamic liquid level is usually detected by the echo meter. This method is limited by the low detection efficiency, discrete measurement, and safety hazards. In addition to this, the existing prediction methods based on data-driven have difficulties in modeling due to the lack of historical data caused by various reasons. In response to the problem about lack of historical data, the multi-scale state characteristics which are closely connected with dynamic liquid level by analyzing the mechanism has been found out in this paper. Further, the state characteristics have been generated by using generative adversarial network (GAN). Simulation experiments have shown that the generated data can be used to establish dynamic liquid level prediction model and solve the problem about lack of historical data. On this basis, support vector regression (SVR) has been adopted as the modeling method and used in oil field production. Finally, the actual prediction results show that the proposed method is effective and meets the application requirements of oilfield engineering.

Key words: dynamic liquid level, soft-sensing, multi-scale state characteristics, generative adversarial networks, support vector regression

中图分类号:

TP 29

侯延彬,高宪文,李翔宇. 采油过程多尺度状态特征生成的有杆泵动态液面预测[J]. 化工学报, 2019, 70(S2): 311-321.

Yanbin HOU,Xianwen GAO,Xiangyu LI. Prediction for dynamic liquid level of sucker rod pumping using generation of multi-scale state characteristics in oil field production[J]. CIESC Journal, 2019, 70(S2): 311-321.

图/表 11

图1 有杆泵抽油系统机理分析

Fig.1 Mechanism analysis of sucker rod pump system

图2 不同沉没度下的示功图

Fig.2 Dynamometer cards in different submergence

图3 生成对抗网络基本流程

Fig.3 Flow chart of generative adversarial networks

图4 基于GAN的采油过程多尺度状态特征生成过程

Fig.4 Generative migration of state characteristics based on GAN

图5 具有较多历史数据的1#、2#抽油井动态液面预测模型仿真图

Fig.5 Prediction for dynamic liquid level of 1# and 2# well with more historical data

表1 具有较多历史数据的各模型的预测精度对比

Table 1 Recognition accuracy prediction of each model with more historical data

建模数据来源	SVR		RBFNN		ELM
建模数据来源	MAPE	RMSE	MAPE	RMSE	MAPE	RMSE
1^#真实数据	1.072%	26.576	1.458%	34.673	1.569%	34.473
1^#生成数据	1.321%	32.640	1.470%	34.839	1.558%	36.298
1^#混合数据	1.209%	28.919	1.295%	34.355	1.610%	44.302
2^#真实数据	3.954%	69.063	8.446%	148.23	10.55%	229.74
2^#生成数据	4.635%	108.10	8.634%	168.50	8.732%	169.64
2^#混合数据	3.985%	73.551	8.093%	154.12	8.459%	199.53

图6 具有较少历史数据的1#、2#抽油井动态液面预测模型仿真图

Fig.6 Prediction for dynamic liquid level of 1# and 2# well with less historical data

表2 具有较少历史数据的各模型的预测精度对比

Table 2 Recognition accuracy prediction of each model with less historical data

建模数据来源	SVR		RBFNN		ELM
建模数据来源	MAPE	RMSE	MAPE	RMSE	MAPE	RMSE
1^#真实数据	1.072%	26.576	1.458%	34.673	1.569%	34.473
1^#混合数据	1.589%	38.954	2.871%	81.781	2.521%	75.997
1^#少量原始数据	1.865%	42.567	3.124%	96.238	2.996%	89.483
2^#真实数据	3.954%	69.063	8.446%	148.23	10.55%	229.74
2^#混合数据	6.923%	156.97	9.439%	166.42	11.53%	259.08
2^#少量原始数据	8.234%	198.86	10.63%	212.35	12.84%	295.34

图7 基于SVR的动态液面预测模型建立流程

Fig.7 Flow chart of prediction for dynamic liquid level based on SVR

图8 油田现场检测设备拓扑、安装示意图

Fig.8 Testing equipment of oil well

表3 油田现场动态液面预测模型精度

Table 3 Recognition accuracy prediction in oil field production

井号

最大相对误差/%

最小相对

误差/%

平均相对

误差/%

均方根

误差

1^#

2^#

3^#

参考文献 31

1	Jia W , Zhou W , Li T F . A review of dynamic fluid level detection for oil well[J]. Applied Mechanics and Materials, 2014, 456: 582-586.
2	Li K , Gao X W , Tian Z D , et al . Using the curve moment and the PSO-SVM method to diagnose downhole conditions of a sucker rod pumping unit[J]. Petroleum Science, 2013, 10(1): 73-80.
3	Li K , Gao X W , Yang W B , et al . Multiple fault diagnosis of down-hole conditions of sucker-rod pumping wells based on Freeman chain code and DCA[J]. Petroleum Science, 2013, 10(3): 347-360.
4	Lieberman S . Automated continuous fluid level monitoring[C]// SPE Production and Operations Symposium. 2005: 69-86.
5	Shao K Y , Yuan M Y , Liu M , et al . Research of oil well dynamic liquid level signal de-noising based on improved wavelet thresholding[J]. Energy Education Science and Technology, 2014, 32(6): 5377-5388.
6	Zhou W , Liu J , Gan L Q . Dynamic liquid level detection method based on resonant frequency difference for oil wells[J]. Turkish Journal of Electrical Engineering and Computer Sciences, 2018, 26(6): 2967-2975.
7	Li P , Cai Y L , Shen X L , et al . An accurate detection for dynamic liquid level based on MIMO ultrasonic transducer array[J]. IEEE Transactions on Instrumentation and Measurement, 2015, 64(3): 582-595.
8	王通, 高宪文, 刘文芳 . 自适应软测量方法在动液面预测中的研究与应用[J]. 化工学报, 2014, 65(12): 4898-4904.
	Wang T , Gao X W , Liu W F . Adaptive soft sensor method and application in determination of dynamic fluid levels[J]. CIESC Journal, 2014, 65(12): 4898-4904.
9	王通, 高宪文, 刘文芳 . 基于改进即时学习算法的动液面软测量建模[J]. 东北大学学报, 2014, 36(7): 918-922.
	Wang T , Gao X W , Liu W F . Soft sensor for determination of dynamic fluid levels based on enhanced just-in-time learning algorithm[J]. Journal of Northeastern University, 2014, 36(7): 918-922.
10	李翔宇, 高宪文, 侯延彬 . 基于在线动态高斯过程回归抽油井动液面软测量建模[J]. 化工学报, 2015, 66(6): 2150-2158.
	Li X Y , Gao X W , Hou Y B . Online dynamic Gaussian process regression for dynamic liquid level soft sensing of sucker-rod pumping well[J]. CIESC Journal, 2015, 66(6): 2150-2158.
11	李翔宇, 高宪文, 李琨, 等 . 基于多源信息特征融合的抽油井动液面集成软测量建模[J]. 化工学报, 2016, 67(6): 2469-2479.
	Li X Y , Gao X W , Li K , et al . Ensemble soft sensor modeling for dynamic liquid level of oil well based on multi-source information feature fusion [J]. CIESC Journal, 2016, 67(6): 2469-2479.
12	Li X Y , Gao X W , Cui Y B , et al . Dynamic liquid level modeling of sucker-rod pumping systems based on Gaussian process regression[C]//International Conference on Natural Computation. 2013: 917-922.
13	Wang G F , Zhou Y W , Fei Y W . Application of BP NN in the measurement of dynamic liquid level[J]. Journal of Test and Measurement Technology,2004, 18(1): 5-8.
14	Bai S , Jiang Z J , Wang T , et al . Application of online SVR on the dynamic liquid level soft sensing[C]//Chinese Control and Decision Conference. 2013: 25-27.
15	Li K , Han Y . Modelling for motor load torque with dynamic load changes of beam pumping units based on a serial hybrid model[J]. Transactions of the Institute of Measurement and Control, 2018, 40(3): 903-917.
16	Li K , Han Y . Soft sensor for dynamic fluid level of beam pump unit based on multiple LS-SVM models[C]//International Conference on Mechatronics and Control. 2014: 2340-2345.
17	Wang T , Lai H Z , Jiang Z J . An algorithm study for determination of dynamic fluid level based on the state space reconstruction and BH-LSSVM[C]//International Conference on Mechatronics and Control. 2014: 132-136.
18	Goodfellow I J , Pouget-Abadie J , Mirza M , et al . Generative adversarial nets[C]//28th Conference on Neural Information Processing Systems. 2014: 2672-2680.
19	Li J , Skinner K , Eustice R , et al . WaterGAN: unsupervised generative network to enable real-time color correction of monocular underwater images[J]. IEEE Robotics and Automation Letters, 2018, 3(1): 387-394.
20	Kumar A , Sattigeri P , Fletcher P T . Semi-supervised learning with GANs manifold invariance with improved inference[C]//31st Conference on Neural Information Processing Systems. 2017.
21	Madani A , Moradi M , Karargyris A . Semi-supervised learning with generative adversarial networks for chest X-ray classification with ability of data domain adaptation[C]// 15th IEEE International Symposium on Biomedical Imaging. 2018.
22	Tu Y , Lin Y , Wang J . Semi-supervised learning with generative adversarial networks on digital signal modulation classification[J]. CMC-Computers Materials & Continua, 2018, 55(2): 243-254.
23	He Z , Liu H , Wang Y W , et al . Generative adversarial networks-based semi-supervised learning for hyperspectral image classification[J]. Remote Sensing, 2017, 9(10):1042.
24	Guo J Y , Lei B , Ding C B , et al . Synthetic aperture radar image synthesis by using generative adversarial nets[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(7): 1111-1115.
25	Yuki S , Shinnosuke T , Hiroshi S . Statistical parametric speech synthesis incorporating generative adversarial networks[J]. IEEE/ACM Transaction on Audio, Speech and Language Processing, 2018, 26(1): 84-96.
26	Xuan Q , Chen Z Z , Liu Y , et al . Multiview generative adversarial network and its application in pearl classification[J]. IEEE Transactions on Industrial Electronics, 2019, 6(10): 8244-8252.
27	Drucker H , Burges C J C , Kaufman L , et al . Support vector regression machines[C]//10th Annual Conference on Neural Information Processing Systems. 1996: 155-161.
28	Bianchini M , Frasconi P , Gori M . Learning without local minima in radial basis function networks[J]. IEEE transaction on Neural Networks,1995, 6(3): 749-756.
29	Huang G B , Zhu Q Y , Siew C K . Extreme learning machine: theory and applications[J]. Neurocomputing, 2006, 70: 489-501.
30	Hou Y B , Gao X W , Wang M S , et al . Application of dynamic liquid level prediction model based on improved SVR in sucker rod pump oil wells[C]//32nd Chinese Control Conference. 2013: 7826-7830.
31	高宪文, 王明顺, 刘占广, 等 . 一种有杆泵抽油井井下动态液位软测量方法: 201310029549[P]. 2015-07-15.
	Gao X W , Wang M S , Liu Z G , et al . Soft measuring method of dynamic liquid level under sucker rod oil well: 201310029549[P]. 2015-07-15.

[1]	闫琳琦, 王振雷. 基于STA-BiLSTM-LightGBM组合模型的多步预测软测量建模[J]. 化工学报, 2023, 74(8): 3407-3418.
[2]	邵伟明, 韩文学, 宋伟, 杨勇, 陈灿, 赵东亚. 基于分布式贝叶斯隐马尔可夫回归的动态软测量建模方法[J]. 化工学报, 2023, 74(6): 2495-2502.
[3]	罗顺桦, 王振雷, 王昕. 基于二子空间协同训练算法的半监督软测量建模[J]. 化工学报, 2022, 73(3): 1270-1279.
[4]	刘聪, 谢莉, 杨慧中. 基于改进DPC的青霉素发酵过程多模型软测量建模[J]. 化工学报, 2021, 72(3): 1606-1615.
[5]	陈忠圣, 朱梅玉, 贺彦林, 徐圆, 朱群雄. 基于分位数回归CGAN的虚拟样本生成方法及其过程建模应用[J]. 化工学报, 2021, 72(3): 1529-1538.
[6]	李东, 黄道平, 刘乙奇. 基于协同训练的半监督异构自适应软测量建模方法的研究[J]. 化工学报, 2020, 71(5): 2128-2138.
[7]	杜宇浩, 阎高伟, 李荣, 王芳. 基于局部线性嵌入的测地线流式核多工况软测量建模方法[J]. 化工学报, 2020, 71(3): 1278-1287.
[8]	杨逸俊,王振雷,王昕. 基于最近邻与神经网络融合模型的软测量建模方法[J]. 化工学报, 2020, 71(12): 5696-5705.
[9]	代学志,熊伟丽. 基于核极限学习机的快速主动学习方法及其软测量应用[J]. 化工学报, 2020, 71(11): 5226-5236.
[10]	廉小亲, 王俐伟, 安飒, 魏伟, 刘载文. 基于SOM-RBF神经网络的COD软测量方法[J]. 化工学报, 2019, 70(9): 3465-3472.
[11]	秦美华, 朱红求, 李勇刚, 陈俊名, 张凤雪, 李文婷. 基于STA-K均值聚类的电化学废水处理过程离子浓度软测量[J]. 化工学报, 2019, 70(9): 3458-3464.
[12]	吴菁, 刘乙奇, 刘坚, 黄道平, 邱禹, 于广平. 基于动态多核相关向量机的软测量建模研究[J]. 化工学报, 2019, 70(4): 1472-1484.
[13]	吉文鹏, 杨慧中. 基于改进扩张搜索聚类算法的多流形软测量建模[J]. 化工学报, 2019, 70(2): 723-729.
[14]	赵荣荣, 赵忠盖, 刘飞. 基于k-近邻互信息的发酵过程高斯过程回归建模[J]. 化工学报, 2019, 70(12): 4741-4748.
[15]	朱宝, 乔俊飞. 基于自编码神经网络特征提取的回声状态网络研究及过程建模应用[J]. 化工学报, 2019, 70(12): 4770-4776.