|本期目录/Table of Contents|

[1]周驰宇,张智,王杨,等.基于神经遗忘决策集成的油管CO2腐蚀速率预测*[J].中国安全生产科学技术,2023,19(6):127-134.[doi:10.11731/j.issn.1673-193x.2023.06.018]
　ZHOU Chiyu,ZHANG Zhi,WANG Yang,et al.Prediction on CO2 corrosion rate of tubing based on neural oblivious decision ensemble[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2023,19(6):127-134.[doi:10.11731/j.issn.1673-193x.2023.06.018]
点击复制

基于神经遗忘决策集成的油管CO2腐蚀速率预测*
分享到：

《中国安全生产科学技术》[ISSN:1673-193X/CN:11-5335/TB]

卷:
19
期数:
2023年6期
页码:
127-134
栏目:
职业安全卫生管理与技术
出版日期:
2023-06-30

文章信息/Info

Title:
Prediction on CO2 corrosion rate of tubing based on neural oblivious decision ensemble
文章编号:
1673-193X(2023)-06-0127-08
作者:
周驰宇; 张智; 王杨; 向世林
（1.西南石油大学 油气藏地质及开发工程国家重点实验室，四川 成都 610500；
2.西南石油大学 计算机科学学院，四川 成都 610500）
Author(s):
ZHOU Chiyu;  ZHANG Zhi;  WANG Yang;  XIANG Shilin
(1.State Key Laboratory of Oil & Gas Reservoir Geology and Exploitation,Southwest Petroleum University,Chengdu Sichuan 610500,China;
2.College of Computer Science,Southwest Petroleum University,Chengdu Sichuan 610500,China)
关键词:
CO2腐蚀速率预测' target="_blank" rel="external">">2腐蚀速率预测; 油管; CO2腐蚀' target="_blank" rel="external">">2腐蚀; 神经网络; 神经遗忘决策集成
Keywords:
CO2 corrosion rate prediction' target="_blank" rel="external">">2 corrosion rate prediction;  tubing;  CO2 corrosion' target="_blank" rel="external">">2 corrosion;  neural network;  neural oblivious decision ensemble
分类号:
TE983X937
DOI:
10.11731/j.issn.1673-193x.2023.06.018
文献标志码:
A
摘要:
为避免CO2腐蚀油管进而破坏井筒完整性，有效预测CO2腐蚀速率，在出现安全隐患之前采取预防和纠正措施。提出1种基于半经验模型预测结果和多维特征构建的数据集训练的神经网络——神经遗忘决策集成(NODE)预测方法，预测某实例井的油管CO2腐蚀速率，并与实测数据和传统模型进行比较，开展敏感因素分析和特征重要性排序。研究结果表明：该方法训练的神经网络预测结果误差较小，可为高效、快速地进行油管CO2腐蚀速率预测和因素分析提供新思路，对井筒完整性设计有一定参考意义。
Abstract:
In order to avoid the CO2 corrosion of tubing and thus damage the wellbore integrity,it is necessary to effectively predict the CO2 corrosion rate,and take the preventive and corrective measures before the emergence of potential safety hazard.A neural network-neural oblivious decision ensemble (NODE) prediction method trained by the dataset constructed on the basis of prediction results of semi-empirical model and multi-dimensional features was put forward,then the CO2 corrosion rate of tubing in an example well was predicted and compared with the measured data and traditional models,and the sensitive factor analysis and feature importance ranking were carried out.The results showed that the error of neural network prediction results trained by this method is small,which can provide a new method for efficient and rapid prediction and factor analysis on the CO2 corrosion rate of tubing,and has certain guiding significance for the design of wellbore integrity.

参考文献/References:

［1］ZHANG N,ZENG D,ZHANG Z,et al.Effect of flow velocity on pipeline steel corrosion behaviour in H2S/CO2 environment with sulphur deposition［J］.Corrosion Engineering,Science and Technology,2018,53(5):370-377.
［2］Norsok Standard No.M506.CO2 corrosion rate calculation model［S］.Norway:Standard Norway,2005.
［3］王献昉,陈长风,白真权,等.CO2腐蚀速率半经验预测模型研究［J］.腐蚀与防护,2009,30(2):90-94.WANG Xianfang,CHEN Changfeng,BAI Zhenquan,et al.Semi-empirical model for prediction of CO2 corrosion rate［J］.Corrosion and Protection,2009,30(2):90-94.
［4］CHENG Y P,LI Z L,LIU Q Q.Research on CO2 corrosion rate prediction of gathering pipelines considering the effect of oil［C］//Applied Mechanics and Materials,2015.
［5］GARBER J D,PATIL V B,YAP K M.CO2,H2S and bacteria pitting corrosion model［C］//NACE International Corrosion 2011 Conference & Expo,2011.
［6］KATTERBAUER K,DOKHON W,AULIA F,et al.A novel corrosion monitoring and prediction system utilizing advanced artificial intelligence［C］//SPE Middle East Oil & Gas Show and Conference,2021.
［7］YANG H,LU L,TSAI K.Machine learning based predictive models for CO2 corrosion in pipelines with various bending angles［C］//SPE Annual Technical Conference and Exhibition,2020.
［8］QU Z,YUE X,LI Y,et al.Corrosion risk prediction based on offshore pipeline service database and machine learning algorithms［C］//AMPP Annual Conference+Expo,2022.
［9］郑度奎,李昊燃,程远鹏,等.油气管道的 CO2腐蚀预测模型和预测方法［J］.腐蚀与防护,2020,41(3):48-53.ZHENG Dukui,LI Haoran,CHENG Yuanpeng,et al.CO2 corrosion prediction model and prediction method for oil and gas pipeline［J］.Corrosion & Protection,2020,41(3):48-53.
［10］张智,刘金铭,张华礼,等.注CO2采油井油管柱腐蚀速率预测［J］.西南石油大学学报(自然科学版),2019,41(2):175-184.ZHANG Zhi,LIU Jinming,ZHANG Huali,et al.Prediction of tubing string corrosion rate in CO2-injection production wells［J］.Journal of Southwest Petroleum University (Science & Technology Edition),2019,41(2):175-184.
［11］谢飞,李佳航,王新强,等.天然气管道CO2腐蚀机理及预测模型研究进展［J］.天然气工业,2021,41(10):109-118.XIE Fei,LI Jiahang,WANG Xinqiang,et al.Research progress on CO2 corrosion mechanism and prediction model of natural gas pipelines［J］.Natural Gas Industry,2021,41(10):109-118.
［12］NESICS,POSTLETHWAITE J,OLSEN S.An electrochemical model for prediction of corrosion of mild steel in aqueous carbon dioxide solutions［J］.Corrosion,1996,52(4):280-294.
［13］DE W C,LOTZ U,MILLIAMS D E.Predictive model for CO2 corrosion engineering in wet natural gas pipelines［J］.Corrosion,1991,47(12):976-985.
［14］POPOV S,MOROZOV S,BAKENKO A.Neural oblivious decision ensembles for deep learning on tabular data［C］//International Conference on Learning Representations,2020.
［15］PETERS B,NICULAE V,MARTINS A F T.Sparse sequence-to-sequence models［C］//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics,2019.

相似文献/References:

[1]陈少波,卢玉斌,袁海梁,等.基于ANSYS/LS-DYNA的液态CO2相变破岩裂纹扩展特征研究*[J].中国安全生产科学技术,2023,19(6):74.[doi:10.11731/j.issn.1673-193x.2023.06.011]
　CHEN Shaobo,LU Yubin,YUAN Hailiang,et al.Study on crack propagation characteristics of liquid CO2 phase transition rock breaking based on ANSYS/LS-DYNA[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2023,19(6):74.[doi:10.11731/j.issn.1673-193x.2023.06.011]

备注/Memo

备注/Memo:
收稿日期： 2023-02-01
* 基金项目： 国家自然科学基金项目（U22A20164，52074234）
作者简介： 周驰宇，硕士研究生，主要研究方向为人工智能。
通信作者： 张智，博士，教授，主要研究方向为井筒完整性与环空带压管控、材料腐蚀与防护。
更新日期/Last Update: 2023-07-09