基于大数据的钻井物料信息多维分析研究

发布时间：2019-01-19 16:07

【摘要】：石油钻井过程中需要使用大量种类和型号繁多的钻具物料,传统钻井物料管理信息系统主要是对钻井物料使用的事务性信息进行管理,缺少对钻井物料使用信息的分析和决策支持,未充分利用和发挥信息化建设所积累的大量数据信息的作用。本文采用大数据的理论和方法,深入研究基于大数据的钻井物料信息多维分析方法,以进一步提高钻井企业管理的效率。第一,构建了基于大数据的钻井物料信息管理系统体系结构。采用大数据理论和方法,结合Hoop技术,以分布式存储的方式,构建基于Hardtop架构的钻井物料信息管理系统的体系结构及进行多维分析的解决方案。第二,提出基于Hive的分布式数据仓库的设计和实现。通过对数据仓库的主题及事实表和维表的设计,构建相应的多维数据模型,并实现数据仓库设计。基于Hive技术设计数据仓库的抽取、转换和加载过程,通过scoop技术实现HDFS与传统关系数据库间数据的相互转移。第三,研究给出基于Hadoop的钻井物料多维分析方法。对所要分析的多维数据进行切片、切块、钻取、旋转等一般操作;并实现基于MapReduce的多维分析,帮助企业决策人员从不同角度分析数据。采用大数据的理论和方法处理钻井物料信息数据,能够充分发挥大数据的成本低、效率高的优势。本研究设计了基于大数据的钻井物料多维分析原型系统,为钻井企业数据分布广、数据量大、难以集成和及时分析处理等问题提供了辅助支持。
[Abstract]:A large number of kinds and types of drilling tool materials are needed in the process of oil drilling. The traditional drilling material management information system mainly manages the transactional information used in drilling materials. There is a lack of analysis and decision support for the information used in drilling materials, and the role of a large amount of data information accumulated in the information construction is not fully utilized and brought into play. In this paper, big data's theory and method are used to deeply study the multidimensional analysis method of drilling material information based on big data, in order to further improve the efficiency of drilling enterprise management. Firstly, the system structure of drilling material information management system based on big data is constructed. By using big data's theory and method, combined with Hoop technology and distributed storage, the system structure of drilling material information management system based on Hardtop architecture and the solution of multidimensional analysis are constructed. Secondly, the design and implementation of distributed data warehouse based on Hive are proposed. Through the design of the subject, fact table and dimension table of the data warehouse, the corresponding multidimensional data model is constructed, and the design of the data warehouse is realized. The extraction, transformation and loading process of data warehouse are designed based on Hive technology, and the data transfer between HDFS and traditional relational database is realized by scoop technology. Thirdly, the multi-dimensional analysis method of drilling materials based on Hadoop is presented. The multidimensional data to be analyzed are sliced, cut, drilled, rotated and so on, and the multidimensional analysis based on MapReduce is realized to help enterprise decision makers to analyze the data from different angles. Using big data's theory and method to deal with drilling material information data can give full play to big data's advantages of low cost and high efficiency. The prototype system of multi-dimensional analysis of drilling materials based on big data is designed in this paper, which provides the auxiliary support for the problems such as wide distribution of drilling data, large amount of data, difficult integration and timely analysis and processing.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TE24;TP311.13

【相似文献】