本文主要研究对象是实时数据库系统中的历史数据库，构建实时数据库系统VegeBam的基本层次结构和研究历史数据库中的多维空间索引技术，并在研究的基础上开发一个具有自主知识产权的实时数据库系统VegeBam。本文的主要内容如下： 描述实时数据库系统VegeBam的需求，并在分析需求的基础上设计VegeBam的存储结构，包括实时数据库系统文件的存储结构和数据表文件的存储结构等。 在多维空间索引结构R-tree的基础上，提出一种改进的多维空间索引结构RS-tree。RS-tree索引结构具有R-tree索引结构所没有的索引记录有序性和节点位置有序性两种特性，这使得在查询过程中RS-tree索引结构的磁盘I/O操作次数比R-tree索引结构明显减少，提升了RS-tree索引结构的查询性能。实验结果表明：在执行连续1 000次以上的查询中，RS-tree索引结构的平均查询速度比R-tree索引结构的平均查询速度要快将近500倍；RS-tree索引结构应对大量查询比R-tree索引结构更具有优势，是一种有效的多维空间索引结构。 设计实时数据库系统VegeBam的层次结构，定义各层之间的接口，并在此基础上开发具有自主知识产权的实时数据库系统VegeBam。在该系统中，数据以二维表的形式存储在数据文件中，并实现历史数据库的创建数据表、增加、删除、修改和查询记录等相关操作。系统内嵌多维数据索引RS-tree，支持多维数据的查询。通过对VegeBam进行数据写入测试、数据读取测试和数据查询测试，结果表明：实时数据库系统VegeBam是一个具有实际应用价值的历史数据库系统。

The main concern in this thesis is the historical database which is a part of Real-time database system. The basic hierarchy architecture of VegeBam is constructed and a dynamic index structure of multi-dimensional space is studied. A Real-time database system called VegeBam is developed, which is of proprietary intellectual property rights. The main contents are as follows. Requirements of VegeBam are described and then storage structures of VegeBam are designed based on requirements including storage structure of the Real-time database file, storage structure of data table file and so on. A dynamic index structure called RS-tree is proposed based on a multi-dimensional index structure called R-tree. RS-tree is of two characteristics, Sequence ordering of index records and Position ordering of nodes, of which R-tree is not. It makes that the number of I/O operations of RS-tree is less than that of R-tree obviously during query, which improves the query performance of RS-tree. Experimental results show that RS-tree's average speed of query is nearly five hundreds of times faster than R-tree's when the number of query records is more than one thousand and when query is performed in huge amount of data, RS-tree is an effective index structure which can improve the query performance obviously. We have not only designed the hierarchy of VegeBam and defined interfaces between hierarchies, but also developed a Real-time database system called VegeBam which is of proprietary intellectual property rights. Data are stored in the form of bivariate table in VegeBam. VegeBam implements several operations including adding records, deleting records, modifying records and querying records, and supports multi-dimensional index structure called RS-tree. We perform several tests including data writing test, data reading test and data querying test. Experimental results show that Real-time database system called VegeBam is of application value fairly.