基于知识图谱和机器学习的油气田地面方案智能平台建设
Development of an intelligent platform for oil and gas field surface development scheme based on knowledge graph and machine learning
浏览(3452) 下载(9)
- 引用格式:
-
宋旭,李宏斌,单吉全,章瑞,王永东,许斌.基于知识图谱和机器学习的油气田地面方案智能平台建设[J].天然气与石油,2025,43(1):30-37.doi:10.3969/j.issn.1006-5539.2025.01.004
SONG Xu, LI Hongbin, SHAN Jiquan, ZHANG Rui, WANG Yongdong, XU Bin.Development of an intelligent platform for oil and gas field surface development scheme based on knowledge graph and machine learning[J].Natural Gas and Oil,2025,43(1):30-37.doi:10.3969/j.issn.1006-5539.2025.01.004
- DOI:
- 10.3969/j.issn.1006-5539.2025.01.004
- 作者:
- 宋旭1 李宏斌2 单吉全3 章瑞3 王永东4 许斌3
SONG Xu1, LI Hongbin2, SHAN Jiquan3, ZHANG Rui3, WANG Yongdong4, XU Bin3
- 作者单位:
- 1. 中国石油集团安全环保技术研究院有限公司, 北京 102200; 2. 大庆油田有限责任公司, 黑龙江 大庆 163712; 3. 中国石油天然气股份有限公司长庆油田分公司, 陕西 西安 710018; 4. 大庆油田设计院有限公司, 黑龙江 大庆 163712
1. CNPC Research Institute of Safety & Environment Technology Co., Ltd., Beijing, 102200, China; 2. Daqing Oilfield Co., Ltd., Daqing, Heilongjiang, 163712, China; 3. PetroChina Changqing Oilfield Company, Xi'an, Shaanxi, 710018, China; 4. Daqing Oilfield Design Institute Co., Ltd., Daqing, Heilongjiang, 163712, China
- 关键词:
- 油气田;地面方案设计;知识图谱;机器学习
Oil and gas fields; Surface development scheme design; Knowledge graph; Machine learning
- 摘要:
油气田地面方案设计涉及的油气田类型、工程类型、设计专业、成果资料类型等众多,方案设计工作专业度高、难度大且复杂,方案设计质量过于依赖个人工作经验,存在方案成果共享、再利用程度低等问题,亟需通过信息化、智能化手段解决。以地面工程知识体系为基础,基于知识图谱和机器学习融合技术构建油气田地面方案智能平台以实现智能检索、智能辅助设计、智能辅助审查等应用场景,自动推荐油气田地面工程项目周边环境、采标、相似工艺方案、审查要点、历史专家意见,自动抽提项目报告中的关键技术指标、经济指标和主要工程量,智能推送对比分析结果等应用。通过在北一区断西东块二类抗盐聚合物产能建设项目方案和龙西地区塔21-4区块产能建设方案的试用验证,油气田地面方案智能平台可节约资料检索耗时,实现“一键即得”,提升自查自审质量、有效减少项目多轮审查频次,实现项目资料在线管理、共享应用,提高设计、审查工作效率超50%,有效提高方案设计审查工作质效。油气田地面方案智能平台可为类似油气田地面建设方案提供参考。
The design of surface development scheme of oil and gas fields involves multiple types of oil and gas fields, projects, engineering disciplines, and deliverable data. The design work for the development scheme is highly specialized, difficult, and complex. The quality of development scheme design relies too much on individual work experience, and there are issues such as low level of sharing and reuse of development scheme results that urgently need to be resolved through information technology and intelligent approaches. Based on the knowledge system of surface facilities, this paper integrates knowledge graphs and machine learning technology to develop intelligent application scenarios: intelligent retrieval, intelligent assisted design, and intelligent assisted review. It is able to generate automatic recommendation of the surrounding environment, standard selection, similar process schemes, review points, and historical expert opinions of surface facilities engineering. It can also automatically extract key technical indicators, economic indicators, and main engineering bill of quantities from project reports, and provides intelligent push of comparative analysis results and other applications. Through the application of the Class Ⅱ salt-resistant polymer production capacity building project plan in the Duanxi East Block of the North-1 District and the production capacity building plan in the Ta21-4 Block of the Longxi District, it is concluded that the intelligent support platform for oil and gas field surface development scheme design can save data retrieval time, achieve “one-click access”, improve the quality of internal-inspection and internalaudit, effectively reduce the frequency of multiple rounds of project review, achieve online management and sharing of project data, increase the work efficiency of design and review personnel by more than 50% and effectively improve the efficiency and quality of development scheme design and review. This intelligent platform can provide reference for similar oil and gas field surface development in the future.