基于井下数据中心虚拟数据整合模型研究

发布时间：2018-03-27 21:05

本文选题：数据虚拟　切入点：井下数据中心　出处：《东北石油大学》2013年硕士论文

【摘要】：随着信息技术的不断发展，，数据种类、数据存储量都在不断增加，企业信息化建设不断加快。由于采用数据标准不统一等原因，企业内部与企业间的数据交流难度越来越大，数据孤岛的数量急剧增加。如何利用好已经获取的数据对企业的业务活动及决策进行有效支撑，已经成为企业信息化过程的一个至关重要的问题。面对着日渐复杂的数据生存环境，建立多个数据孤岛之间的交流已是企业信息化建设的首要任务。通过分析井下作业分公司数据中心建设中各数据源存在的分布性、异构性、自治性等特点。对虚拟数据整合模型及其实现技术进行充分研究，提出了虚拟数据整合体系结构，构建了虚拟数据整合元数据模型，基于该模型实现信息孤岛间的数据整合。首先，针对数据源分散状况，建立了数据源层。利用元数据技术，建立了虚拟数据整合数据场地元模型。通过元模型实现多个分散数据源的逻辑整合，形成数据统一访问逻辑接口。为有效解决数据源访问效率低的问题，采用数据分割技术构建数据访问层。细化了所访问数据的粒度，提高数据访问效率。其次，针对异构数据模型间的整合问题，以数据中心数据字典为框架，建立了对多个异构数据模型的映射支撑模型。通过元数据模型驱动实现数据转换层，对分割后的数据实现数据转换。同时结合内存映射原理，引用MongoDB作为数据缓存带，将读取到的大量数据存入MongoDB缓存，对缓存的数据实现并行转换处理，提高了数据整合的效率。最后，基于虚拟数据整合元模型，结合中间件、数据物化与NOSQL相关技术，实现数据应用层，参照虚拟数据整合模型中应用元模型对数据进行整合，展示、转储，满足企业对数据的应用需求。本文通过研究虚拟整合模型及其实现技术研究，实现基于元数据的虚拟数据整合体系框架，并实现了原型系统，验证了虚拟整合体系结构合理性，完成井下数据中心虚拟数据整合，实践证明模型及技术的可行性。
[Abstract]:With the continuous development of information technology, the data types and data storage are increasing, and the construction of enterprise information is accelerating. Because of different data standards, it is more and more difficult to communicate data between enterprises and enterprises. The number of data islands has increased dramatically. How to make good use of the acquired data to effectively support the business activities and decisions of enterprises, It has become a crucial problem in the process of enterprise informatization. In the face of increasingly complex data living environment, It is the most important task of enterprise information construction to establish the exchange among multiple data islands. By analyzing the distribution and heterogeneity of the data sources in the construction of data center of underground operation branch, the paper analyzes the distribution and heterogeneity of the data sources in the construction of the data center of the underground branch. Based on the study of the virtual data integration model and its implementation technology, the virtual data integration architecture is proposed, and the virtual data integration metadata model is constructed. Based on the model, the data integration between isolated information islands is realized. Firstly, according to the situation of data source dispersion, the data source layer is established. Using metadata technology, the virtual data integration data site meta-model is established. Through the meta-model, the logical integration of multiple decentralized data sources is realized. In order to solve the problem of low access efficiency of data sources, the data access layer is constructed by using data segmentation technology, which refines the granularity of the data accessed and improves the efficiency of data access. Secondly, aiming at the problem of integration between heterogeneous data models, a mapping support model of multiple heterogeneous data models is established based on the data center data dictionary, which is driven by metadata model to realize data conversion layer. At the same time, the MongoDB is used as the data cache band, the large amount of data is stored in the MongoDB cache, the cached data is converted into parallel processing, and the efficiency of data integration is improved. Finally, based on virtual data integration metamodel, combining middleware, data materialization and NOSQL technology, the data application layer is realized. According to the application metadata model of virtual data integration model, the data is integrated, displayed and stored. To meet the data application requirements of enterprises. Through the research of virtual integration model and its implementation technology, this paper realizes the framework of virtual data integration based on metadata, and implements the prototype system, which verifies the rationality of virtual integration architecture. The virtual data integration of the underground data center is completed, and the feasibility of the model and technology is proved by practice.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP308

【参考文献】