利用优化BP神经网络建立裂缝宽度预测模型

何涛; 谢显涛; 王君; 赵洋; 苏俊霖

doi:10.3969/j.issn.1001-5620.2021.02.012

利用优化BP神经网络建立裂缝宽度预测模型

doi: 10.3969/j.issn.1001-5620.2021.02.012

何涛¹,
谢显涛¹,
王君¹,
赵洋²,
苏俊霖^2,3

1. 中国石油天然气集团川庆钻探工程有限公司, 成都 610000;
2. 西南石油大学油气藏地质及开发工程国家重点实验室, 成都 610000;
3. 西南石油大学石油与天然气工程学院, 成都 610000

基金项目:

川庆钻探-西南石油大学创新联合体科技合作项目“川渝地区钻井防漏堵漏数据挖掘技术研究与应用”（CQXN-2019-06）；中国石油-西南石油大学创新联合体科技合作项目（2020CX040201）

详细信息

作者简介:
何涛,1979年生,高级工程师,现在从事钻井液技术研究和现场服务工作

中图分类号: TE282
计量
- 文章访问数: 567
- HTML全文浏览量: 169
- PDF下载量: 46
- 被引次数: 0
出版历程
- 收稿日期: 2020-12-12

Crack Width Prediction Model Based On Optimized BP Neural Network

1. China National Petroleum Corporation Chuanqing Drilling Engineer Co., Ltd., Drilling Fluid Technology Service Company, Chengdu, Sichuan 610000;
2. State Key Laboratory of Oil and Gas Reservoir Geology and Development, Southwest Petroleum University, Chengdu, Sichuan 610000;
3. Institute of Petroleum and Natural Gas Engineering, Southwest Petroleum University, Chengdu, Sichuan 610000

摘要

摘要: 裂缝性井漏问题严重威胁着钻井生产安全与进度，目前现场堵漏成功率较低，其中一个重要原因就是无法准确预测裂缝宽度，裂缝宽度认知的局限性导致难以确定堵漏的方法及材料颗粒大小。因此本文提出了利用优化BP神经网络方法建立井漏裂缝宽度预测模型，用以解决裂缝宽度预测难的问题。首先根据方差分析（ANOVA）法确定了影响裂缝宽度的相关参数，将其输入优化的BP神经网络模型进行训练，并通过样本数据拟合验证预测精度，最终测试集数据拟合直线的斜率为0.8772，截距为0.0206。另外，为了确认模型稳定性，针对裂缝宽度预测模型进行了性能评估，得出该模型确定系数（R²） 0.89，平均绝对百分比误差（PCC） 0.82，均方根误差（RMSE） 1.35，证明该模型性能优良。最后利用现场工程数据进行进行实例预测，由结果可知，该模型具有较高的预测精度，可以在堵漏工程作业中提供较好的辅助决策。
- 井漏 /
- BP神经网络 /
- 裂缝宽度 /
- 拟牛顿法
Abstract: The problem of fractured well loss is a serious threat to the safety and progress of drilling production. At present, the success rate of field plugging is relatively low one of the important reasons is that the fracture width cannot be accurately predicted, and the limitations of fracture width cognition make it difficult to determine the plugging method and material particle size. Therefore this paper puts forward to use the improved BP neural network method to establish the prediction model of lost circulation fracture width in order to solve the problem of fracture width prediction. Firstly, according to the analysis of variance method, the related parameters that affect the crack width are determined which are input into the improved BP neural network model for training, and the accuracy is verified by the sample data prediction. Finally, the slope of the fitting straight line of the test set data is 0.8772, and the intercept is 0.0206. In addition, in order to confirm the stability of the model, the performance of the crack width prediction model is evaluated, and the model's determination coefficient (R²) is 0.89, the average absolute percentage error (PCC) is 0.82, and the root mean square error (RMSE) is 1.35, which proves that the performance of the model is excellent. Finally, the field engineering data is used for example prediction. From the results, it can be seen that the model has high prediction accuracy and can provide better assistant decision-making in plugging engineering operation.
- Well leakage /
- BP neural network /
- Crack width /
- L-bfgs method

HTML全文

参考文献(12)

[1]	徐鹏. 斜井漏失机理及堵漏效果评价实验[D]. 中国石油大学(华东), 2016. XU Peng. Leakage loss mechanism and evaluation of plugging effect[D]. China University of Petroleum, 2016.
[2]	何龙, 史堃, 杨健, 等. 裂缝性地层堵漏材料承压性能及材料优选研究[J]. 钻采工艺, 2019, 42(2):42-44. HE Long, SHI Wei, YANG Jian, et al. Study on pressure bearing performance and material optimization of fractured formations in fractured strata[J]. Drilling & Production Technology, 2019, 42(2):42-44.
[3]	陈阳. 天然裂缝有效性影响控制因素浅析[J]. 石油化工应用, 2016, 35(3):57-59. CHEN Yang. Analysis on the control factors of the effectiveness of natural cracks[J].Petrochemical Industry Application, 2016, 35(3):57-59.
[4]	赵洋, 邓明毅, 曾文强, 等.Griffiths天然裂缝宽度预测模型研究与分析[J]. 钻采工艺, 2017, 40(5):102-105. ZHAO Yang, DENG Mingyi, ZENG Wenqiang, et al. Research and analysis of Griffiths natural fracture width prediction model[J]. Drilling and Production Technology, 2017, 40(5):102-105.
[5]	彭浩, 李黔, 尹虎, 等.Lietard天然裂缝宽度预测模型求解新方法[J]. 石油钻探技术, 2016, 44(3):72-76. PENG Hao, LI Qian, YIN Hu, et al. A new method for solving lietard natural fracture width prediction model[J]. Petroleum Drilling Technology, 2016, 44(3):72-76.
[6]	陈曾伟. 基于神经网络算法的井下裂缝诊断与堵漏技术[J]. 钻井液与完井液, 2019, 36(1):20-24. CHEN Zengwei. Diagnosis and plugging technology of underground crack based on neural network algorithm[J]. Drilling Fluid & Completion Fluid, 2019, 36(1):20-24.
[7]	唐惠康, 李翔宇.BP神经网络用于某油田的生产动态分析[J]. 油气田地面工程, 2015, 34(4):20-21. TANG Huikang,LI Xiangyu. BP neural netwo-rk applied to the production performance ana-lysis of an oilfield[J].Oil and Gas Field Sur-Face Engineering, 2015, 34(4):20-21.
[8]	郭文霞. 硬地网球男子单打比赛战术诊断指标体系构建与应用[D]. 北京体育大学, 2018. GUO Wenxia. Construction and application of tactical diagnostic index system for men's si-ngles in hard ground tennis[D]. Beijing Spo-rt University, 2018.
[9]	罗悦, 林冰, 温川飙. 基于神经网络的中医体质与体检指标关联模型算法研究[J]. 时珍国医国药, 2018, 29(3):763-766. LUO Yue, LIN Bing, WEN chuanbiao. Study o-n the algorithm of the correlation model bet-ween the constitution and physical examinati-on index of traditional Chinese medicine bas-ed on neural network[J].Shi Zhen Guo Yi G-uo Yao, 2018, 29(3):763-766.
[10]	廖藤藤, 李国庆, 李若溪. 基于带搜索约束的神经网络实现吸收稳定系统操作优化[J]. 石油学报(石油加工), 2017, 33(6):1138-1145. LIAO tengtengteng, LI Guoqing, LI Ruoxi. O-ptimization of operation of absorption stabilit-y system based on neural network with sear-ch constraint[J].Journal of Petroleum (petro-leum processing), 2017, 33(6):1138-1145.
[11]	刘鲭洁, 陈桂明, 刘小方, 等. BP神经网络权重和阈值初始化方法研究[J]. 西南师范大学学报(自然科学版), 2010, 35(6):137-141. LIU Yujie, CHEN Guiming, LIU Xiaofang, et al.Study on BP neural network weight and t-hreshold initialization method[J].Journal of S-outhwest China Normal University(Natural S-cience), 2010, 35(6):137-141.
[12]	Ahmed K. Abbas, Ali A. Bashikh, Hayder Ab-bas, et al. Intelligent decisio-ns to stop or mitigate lost circulation based on machine learning[J]. Energy, 2019, 183.