基于Parzen估计器增强贝叶斯算法的深水表层导管喷射安装参数综合优化框架

doi:10.7623/syxb202603010

石油学报 ›› 2026, Vol. 47 ›› Issue (3): 641-658.DOI: 10.7623/syxb202603010

• 石油工程 • 上一篇

基于Parzen估计器增强贝叶斯算法的深水表层导管喷射安装参数综合优化框架

宋宇¹, 杨进², 宋泽华¹

1. 中国石油大学(北京)人工智能学院北京 102249;
2. 中国石油大学(北京)安全与海洋工程学院北京 102249

收稿日期:2025-01-14 修回日期:2026-01-10 发布日期:2026-04-09
通讯作者: 宋宇,男,1989年11月生,2019年获中国石油大学(北京)博士学位,现为中国石油大学(北京)自动化系副教授,主要从事海洋油气钻井、钻井智能控制及大数据优化决策等研究工作。Email:songyu_cup@163.com
作者简介:宋宇,男,1989年11月生,2019年获中国石油大学(北京)博士学位,现为中国石油大学(北京)自动化系副教授,主要从事海洋油气钻井、钻井智能控制及大数据优化决策等研究工作。Email:songyu_cup@163.com
基金资助:
国家自然科学基金青年科学基金项目(No.5220040087)和国家重点研发计划项目(2022YFC28061004)资助。

Comprehensive optimization framework for deepwater surface conductor jetting parameters based on an enhanced Bayesian algorithm with Parzen estimator

Song Yu¹, Yang Jin², Song Zehua¹

1. College of Artificial Intelligence, China University of Petroleum, Beijing 102249, China;
2. College of Safety and Ocean Engineering, China University of Petroleum, Beijing 102249, China

Received:2025-01-14 Revised:2026-01-10 Published:2026-04-09

摘要/Abstract

摘要： 深水表层导管喷射安装是海洋油气开发的关键环节,其作业时效与井口稳定性直接制约着深水建井的工程周期与经济成本。导管喷射安装时间由喷射钻进时间和静置恢复时间两部分组成,二者受多种喷射参数的共同影响,其复杂性及相互制约关系使得传统优化方法难以实现。采用正交模拟实验数据,同时考虑喷射钻进与静置两阶段时间的复杂关联,以及不同地层和操作参数之间的非线性影响,开发出并行预测喷射钻进时间与静置恢复时间的多输出预测模型。通过转化所建立的多输出预测模型,设计基于树结构Parzen估计器与高斯过程相结合的分阶段贝叶斯优化方法应用于综合目标优化,进而调整优化目标权重系数,揭示出优化参数在不同时间阶段效率提升上的关键作用。以深度分别为70 m、80 m、90 m的目标井为案例进行了实证分析。建立了最优的喷射钻井参数组合,新方法较常规设计方法喷射钻进时间平均增加了2.91 %,静置恢复时间平均减少了14.98 %,总作业安装时间减少了6.8 % 。通过量化多维参数间的非线性制约关系,实现了作业时效与井口稳定性的全局协同优化,为喷射参数精细化决策提供了算法指导。

关键词: 深水表层导管喷射安装, 喷射安装参数优化, 贝叶斯优化算法, 模拟正交试验, 并行预测

Abstract: T he jetting installation of deepwater surface conductors is a pivotal process in offshore oil and gas development. Its operational efficiency and wellhead stability directly influence the engineering timeline and economic costs associated with deepwater well construction. The total installation time is comprised of two phases:jetting drilling time and shut-in time. Both phases are influenced by a range of jetting parameters. Due to the complexity and interdependencies among these parameters ose significant challenges for conventional optimization methods in achieving satisfactory solutions. This study employs data from orthogonal simulation experiments, taking into account the intricate relationships between jetting drilling time and shut-in time, as well as nonlinear interactions among distinct geological formations and operational parameters. On this basis, a multi-output predictive model was developed to simultaneously forecast both jetting drilling time and shut-in time. Building upon this model, a staged Bayesian optimization method was devised, integrating the Tree-structured Parzen Estimator with Gaussian Processes for comprehensive objective optimization. By adjusting the weight coefficients of the optimization targets, this approach reveals the critical role of optimization parameters in improving efficiency at different stages of the process. Empirical analysis was conducted on target wells with depths of 70 m, 80 m, and 90 m. An optimal combination of jetting parameters was identified, resulting in a 2.91 % increase in jetting drilling time on average, compared to conventional design methods. However, the shut-in time was reduced by 14.98 %, leading to a 6.8 % reduction in total installation time. By quantifying the nonlinear interdependencies among multi-dimensional parameters, this research achieves a global collaborative optimization of operational efficiency and wellhead stability. This approach offers algorithmic guidance for the refined decision-making of jetting parameters.

Key words: deepwater surface conductor jetting installation, optimization of jetting installation parameters, Bayesian optimization algorithm, orthogonal simulation experiment, parallel predictive model

中图分类号:

TE928

宋宇, 杨进, 宋泽华. 基于Parzen估计器增强贝叶斯算法的深水表层导管喷射安装参数综合优化框架[J]. 石油学报, 2026, 47(3): 641-658.

Song Yu, Yang Jin, Song Zehua. Comprehensive optimization framework for deepwater surface conductor jetting parameters based on an enhanced Bayesian algorithm with Parzen estimator[J]. Acta Petrolei Sinica, 2026, 47(3): 641-658.

参考文献

[1] 杨进,傅超,刘书杰,等.超深水浅层建井关键技术创新与实践[J].石油学报,2022,43(10):1500-1508.
YANG Jin,FU Chao,LIU Shujie,et al.Key technological innovation and practice of well construction in ultra-deepwater shallow formations[J].Acta Petrolei Sinica,2022,43(10):1500-1508.
[2] 杨进,李磊,宋宇,等.中国海洋油气钻井技术发展现状及展望[J].石油学报,2023,44(12):2308-2318.
YANG Jin,LI Lei,SONG Yu,et al.Current status and prospects of offshore oil and gas drilling technology development in China[J].Acta Petrolei Sinica,2023,44(12):2308-2318.
[3] YANG Jin.Jetting installation method of oil-gas well conductor[M]//YANG Jin.Installation methods of offshore oil-gas well conductor.Singapore:Springer,2023:107-162.
[4] 孙宝江,周梓强,孙骞.油气钻采工程中的人工智能方法研究进展[J].石油学报,2025,46(11):2141-2173.
SUN Baojiang,ZHOU Ziqiang,SUN Qian.Progress on artificial intelligence methods in oil and gas drilling and production[J].Acta Petrolei Sinica,2025,46(11):2141-2173.
[5] 刘书杰,杨进,周建良,等.深水海底浅层喷射钻进过程中钻压与钻速关系[J].石油钻采工艺,2011,33(1):12-15.
LIU Shujie,YANG Jin,ZHOU Jianliang,et al.Research on relationship between weight-on-bit and drilling rate during jetting drilling in sub-bottom deepwater[J].Oil Drilling & Production Technology,2011,33(1):12-15.
[6] KAN Changbin,YANG Jin,YU Xiaocong,et al.Field experimental investigation of bit stick-out for different soil strengths during deepwater conductor injection[J].Journal of Petroleum Science and Engineering,2018,169:825-836.
[7] 杨进,严德,田瑞瑞,等.深水喷射下表层导管合理钻头伸出量计算[J].石油勘探与开发,2013,40(3):367-370.
YANG Jin,YAN De,TIAN Ruirui,et al.Bit stick-out calculation for the deepwater conductor jetting technique[J].Petroleum Exploration and Development,2013,40(3):367-370.
[8] 李磊,赵颖,周波,等.深水喷射钻井等候时间对表层导管稳定性影响分析及设计方法[J].石油钻采工艺,2025,47(6):674-683.
LI Lei,ZHAO Ying,ZHOU Bo,et al.Analytic approach for the influence of waiting time on deepwater conductor stability[J].Oil Drilling & Production Technology,2025,47(6):674-683.
[9] 许云锦,杨进,孟炜,等.深水钻井导管喷射过程中钻头与海底土的相互作用[J].石油钻采工艺,2015,37(1):47-49.
XU Yunjin,YANG Jin,MENG Wei,et al.Interaction between drill bit and subsea soil during deepwater drilling pipe ejection[J].Oil Drilling & Production Technology,2015,37(1):47-49.
[10] 周波,杨进,周建良,等.深水钻井喷射下导管排量设计方法[J].石油钻探技术,2016,44(3):21-26.
ZHOU Bo,YANG Jin,ZHOU Jianliang,et al.A jetting flow rate design method for conductor installation through jetting in deepwater drilling[J].Petroleum Drilling Techniques,2016,44(3):21-26.
[11] 王凌霄.深水钻井喷射下导管技术水力参数优化[D].青岛:中国石油大学(华东),2016.
WANG Lingxiao.Hydraulic parameters optimization of conductor jetting technology in deepwater drilling[D].Qingdao:China University of Petroleum (East China),2016.
[12] 刘和兴,方满宗,刘智勤,等.南海西部陵水区块超深水井喷射下导管技术[J].石油钻探技术,2017,45(1):10-16.
LIU Hexing,FANG Manzong,LIU Zhiqin,et al.Jetting-based conductor running technology used in ultra-deep water well of Lingshui block in the western South China Sea[J].Petroleum Drilling Techniques,2017,45(1):10-16.
[13] 阚长宾.深水钻井导管高效喷射安装机理及安全控制研究[D].北京:中国石油大学(北京),2018.
KAN Changbin.High efficiency jetting mechanism and safety control research on deepwater conductor installation[D].Beijing:China University of Petroleum,2018.
[14] WANG Teng,SONG Bin.Study on deepwater conductor jet excavation mechanism in cohesive soil[J].Applied Ocean Research,2019,82:225-235.
[15] FU Chao,YANG Jin,LONG Yang,et al.Calculation method of deep-water surface conductor bearing capacity based on jetting operation parameters[J].Geoenergy Science and Engineering,2024,234:212625.
[16] PELIKAN M.Bayesian optimization algorithm:from single level to hierarchy[D].Champaign:University of Illinois at Urbana-Champaign,2002.
[17] P ELIKAN M,GOLDBERG D E,CANTU ＇ -PAZ E.BOA:the Bayesian optimization algorithm[C] //Proceedings of the 1st Annual Conference on Genetic and Evolutionary Computation.Orlando:Morgan Kaufmann Publishers Inc.,1999:525-532.
[18] 杨进,张明贺,张伟国,等.深水浅层钻井"三合一"高效作业模式[J].石油学报,2025,46(3):599-608.
YANG Jin,ZHANG Minghe,ZHANG Weiguo,et al."Three-in-one" efficient drilling operation mode in deepwater shallow Formations[J].Acta Petrolei Sinica,2025,46(3):599-608.
[19] FANG Tiancheng,REN Fushen,LIU Hanxu,et al.Progress and development of particle jet drilling speed-increasing technology and rock -breaking mechanism for deep well[J].Journal of Petroleum Exploration and Production Technology,2022,12(6):1697-1708.
[20] LI Huan,LI Jingbin,HUANG Zhongwei,et al.Experimental investigation into the effect of process parameters on the drilling performance for swirling impeller abrasive water jet[J].Rock Mechanics and Rock Engineering,2023,56(12):8833-8850.
[21] R AMBA V,SELVARAJU S,SUBBIAH S,et al.Optimization of drilling parameters using improved play-back methodology[J].Journal of Petrol eum Science and Engineering,2021,206:108991.
[22] RANDOLPH M,GOURVENEC S.Offshore geotechnical engineering[M].London:CRC Press,2017.
[23] SONG Yu,SONG Zehua,YANG Jin,et al.Multi-objective optimization framework for deepwater riser jetting installation parameters using deep reinforcement learning[J].Ocean Engineering,2024,309:118398.
[24] J OSEPH V R.Optimal ratio for data splitting[J].Statistical Analysis and Data Mining,2022,15(4):531-538.
[25] BERGSTRA J,BARDENET R,BENGIO Y,et al.Algorithms for hyper-parameter optimization[C]//Proceedings of the 25th International Conference on Neural Information Processing Systems.Granada:Curran Associates Inc.,2011:2546-2554.
[26] S NOEK J,LAROCHELLE H,ADAMS R P.Practical Bayesian optimization of machine learning algorithms[C]//Proceedings of the 26th Inte rnational Conference on Neural Information Processing Systems.Lake Tahoe:Curran Associates Inc.,2012:2951-2959.
[27] B ERGSTRA J,KOMER B,ELIASMITH C,et al.Hyperopt:a python library for model selection and hyperparameter optimization[J].Computatio nal Science & Discovery,2015,8(1):014008.
[28] GARRO B A,VÁZQUEZ R A.Designing artificial neural networks using particle swarm optimization algorithms[J].Computational Intelligence and Neuroscience,2015,2015:369298.
[29] HE Kaiming,ZHANG Xiangyu,REN Shaoqing,et al.Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:770-778.
[30] LOSHCHILOV I,HUTTER F.SGDR:stochastic gradient descent with warm restarts[C]//Proceedings of the 5th International Conference on Learning Representations.Toulon:OpenReview.net,2017.
[31] BAO Yukun,LIU Zhitao.A fast grid search method in support vector regression forecasting time series[C]//Proceedings of the 7th International Conference on Intelligent Data Engineering and Automated Learning.Burgos:Springer,2006:504-511.
[32] P RECHELT L.Early stopping—but when?[M]//MONTAVON G,ORR G B,MÜLLER K R.Neural networks:tricks of the trade.2nd ed.Berl in,Heidelberg:Springer,2012:53-67.
[33] KAMATH C.Intelligent sampling for surrogate modeling,hyperparameter optimization,and data analysis[J].Machine Learning with Applications,2022,9:100373.

基于Parzen估计器增强贝叶斯算法的深水表层导管喷射安装参数综合优化框架

Comprehensive optimization framework for deepwater surface conductor jetting parameters based on an enhanced Bayesian algorithm with Parzen estimator

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 7

编辑推荐

Metrics

本文评价

[1]	徐宝昌, 张学智, 王雅欣, 刘伟, 孟卓然. 用于两相流环空压力预测的自适应物理信息神经网络模型[J]. 石油学报, 2023, 44(3): 545-555.
[2]	汪跃龙, 李凌云, 贺艳, 毛艳慧, 高怡. 近钻头钻具姿态测量的多传感器最小二乘原理加权融合方法[J]. 石油学报, 2021, 42(4): 500-507.
[3]	徐宝昌, 孟宇, 刘伟. 控压钻井井下不可测变量的非线性估计[J]. 石油学报, 2016, 37(12): 1543-1549.
[4]	宗艳波, 王磊, 史晓锋. SAGD成对水平井磁导向测距仪研制与应用[J]. 石油学报, 2016, 37(11): 1428-1434.
[5]	汤楠, 霍爱清, 汪跃龙, 程为彬. 旋转导向钻井稳定平台控制对象动态特性研究[J]. 石油学报, 2009, 30(4): 598-602.
[6]	汤楠, 霍爱清, 汪跃龙, 程为彬. 旋转导向钻井工具稳定平台控制功能试验研究[J]. 石油学报, 2008, 29(2): 284-287.
[7]	杨志龙. 石油钻进过程最优控制[J]. 石油学报, 2000, 21(5): 77-85.