Research And Development On History Database In Process Companies MIS

There is a great demand of dealing with real-time data in process manufacturing companies, and being able to analyze and manage real-time data is the requirement of management automation and decision-making. This article analyzes the application requirement, introduces the structure of a real-time supervision system--Beehive, and then describes some key technologies with history data processing.This article studies two main thesises, one is the design of history data server based on SQL Server 2000, and the other is the design of some software tools used to retrieve data from history tables.The proper design of database is the most important part of this project, which is the base of the framework of system configuration, the development of software tools and the design of application. Our system is designed with Microsoft SQL Server2000. Based on the analysis of the theories concerned with the development, including the basic theory about relation database, the operation data and the analysis data, and the characteristic and optimization of SQL Server2000, this paper presents a database design suitable for the real-time supervision system, which is made up of the set up of the new database, management and performance of tasks.Two main contents of designing the database are proposed, one is to retrieve zipped data according to some rules, which is fulfilled by a few procedures together and can be controlled by input parameters. Anther is to clear up the real-time data stored in database timely, which is fulfilled by using DTS function in SQL Server 2000, and the data out of time can be deleted or saved as a Access file.Meanwhile, based on the analysis of the mechanism and framework of the Beehive system, comparing with other similar system, the programming and testing of two software tools querying history data are finished, including EZSeek, EZReport and database procedures. EZSeek and EZReport introduce some techniques, such as general data querying, and OLE automation to process data and make reports. These software tools are constructed on the document/view architecture, which can display data trend in three views.Finally, embarrassed by the complex application of reports, I realized that it is not enough only constructing and administrating a data warehouse. After making an investigation about the theories of data warehouse, a data mart is proposed which obtains report data using two time tasks, and further suggestions about system expansion are set forth.