基于HBase的医疗卫生数据中心构建与异构数据库同步研究

发布时间：2018-05-19 23:02

本文选题：医疗卫生 + HBase　；参考：《电子科技大学》2013年硕士论文

【摘要】：随着社会信息化的飞速发展，，国家卫生部就推进医疗卫生行业信息化建设作出专门的强调。本文则以省卫生厅的“区域医疗监管平台”课题为研究基础，针对医疗卫生行业的信息孤岛、信息协同共享及监管困难等难题，提出构建医疗卫生信息数据中心，并把所有医疗行业的数据同步到数据中心的设计思想来解决这些难题。在此过程中就需要解决医疗卫生数据中心的构建、异构关系数据库到数据中心的同步和数据中心效率问题就构成本文研究内容。通过深入的研究分析医疗卫生信息数据中心规模及相关需求问题。在对比和分析关系数据库与新型的NoSQL数据库的差异后提出运用NoSQL数据库来构建，通过分析几种NoSQL数据库特点确定以HBase构建医疗卫生信息数据中心。为确定数据模型，提出以关系数据库的E-R建模基础提出HBase的C-O-R建模思想，然后综合分析卫生部颁布医疗卫生元数据标准和部分医疗卫生机构的实际情况，实现医疗卫生信息的HBase数据中心构建。为实现把各医疗机构的异构关系数据库中的数据透明无差异的同步到数据中心，提出以下方案。首先，在数据格式上提出异构数据库产生的异构数据进行标准化的XML和JSON通用数据格式封装实现屏蔽异构数据库的数据差异；其次是传输协议简单及通用，采用SOA架构设计思想，提出运用Web Service方法实现数据同步传输；在异构数据库的增量数据获取方面揉合时间戳、触发器及日志法为一体的方法；最后，提出通用前置机设计实现对所有异构数据库的读取，实现采用XML文件对异构数据库的差异配置。在构建好HBase中心并拥有数据后，由于HBase对于数据访问只有Row Key的定位及全表扫描两种，为了提高复杂查找效率而提出HBase的列索引构建。本文提出两种索引设计方式：运用Row Key优势和构建索引表。本文索引表的构建是采用MySQL数据库与HBase数据库相结合的双索引体系结构。最后经过模拟测试分析，并同MySQL的相关测试性能进行对比，验证出本文的设计方法可以较好的完成医疗卫生信息中心的构建和异构数据库到数据中心的同步，实现医疗卫生数据共享、监管及业务协同。
[Abstract]:With the rapid development of social informatization, the Ministry of Health has made a special emphasis on promoting the construction of medical and health industry informatization. On the basis of the research of "Regional Medical Supervision platform" of the provincial health department, aiming at the problems of information isolated island, information sharing and supervision difficulties in the medical and health industry, this paper puts forward the construction of the medical and health information data center. And all medical industry data synchronization to the data center design ideas to solve these problems. In this process, it is necessary to solve the construction of medical and health data center. The synchronization between heterogeneous relational database and data center and the efficiency of data center constitute the content of this paper. Through in-depth research and analysis of medical and health information data center size and related needs. After comparing and analyzing the difference between the relational database and the new NoSQL database, the author put forward to use the NoSQL database to construct the medical and health information data center by analyzing the characteristics of several NoSQL databases. In order to determine the data model, this paper puts forward the C-O-R modeling idea of HBase based on E-R modeling of relational database, and then synthetically analyzes the actual situation of medical and health metadata standard promulgated by the Ministry of Health and some medical and health institutions. Construction of HBase data center for medical and health information. In order to synchronize the data in heterogeneous relational database of medical institutions to the data center transparently, the following scheme is put forward. First of all, in the data format, the paper puts forward the standard XML and JSON universal data format encapsulation of heterogeneous data generated by heterogeneous database to mask the data difference of heterogeneous database, and secondly, the transmission protocol is simple and universal. Adopting the idea of SOA architecture design, this paper puts forward the method of synchronizing data transmission using Web Service method, combining time stamp, trigger and log method in the incremental data acquisition of heterogeneous database. Finally, In this paper, a universal front computer is designed to read all heterogeneous databases, and the differential configuration of heterogeneous databases is realized by using XML file. After constructing the HBase center and having the data, the column index construction of HBase is proposed in order to improve the efficiency of complex lookup because HBase has only Row Key positioning and full table scanning for data access. This paper presents two methods of index design: using the advantage of Row Key and constructing index table. In this paper, the construction of index table is based on the combination of MySQL database and HBase database. Finally, through the simulation test analysis, and compared with the related test performance of MySQL, it is verified that the design method of this paper can better complete the construction of medical and health information center and the synchronization of heterogeneous database to data center. Achieve medical and health data sharing, supervision and business collaboration.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13;TP308

【参考文献】