【作者】 钟平；

【导师】 李舟军；

【作者基本信息】 国防科学技术大学 ， 软件工程， 2006， 硕士

【摘要】 火力发电厂实时数据库系统存储了大量DCS(Distributed Control System)系统的数据,这些数据的背后隐藏着许多有益于提高火力发电厂运行效率和安全的信息。但由于没有被深刻理解和挖掘,不可避免地造成了数据资源的巨大浪费。数据挖掘技术能够从数据中自动地提取知识,本文致力于研究数据挖掘技术在火力发电厂中的应用,以充分发挥存储在数据库中大量DCS数据对电力生产的指导作用。本文围绕数据挖掘技术,考察和探讨了数据挖掘技术在火力发电厂中的应用现状,并主要研究了数据挖掘技术中的数据预处理技术、模糊C均值聚类算法和OLAP等技术及其应用。针对火力发电厂的数据特性,本文尝试将数据预处理技术应用到实时数据库的数据处理中去,使数据的有效性得到保证,为数据挖掘的进一步工作打下基础。本文在对凝汽器性能分析的基础上,尝试构建了凝汽器多维数据模型,并应用模糊C均值聚类算法对多维数据集进行处理。在此基础上,在Excel的数据透视表中进行了以凝汽器真空为主题的OLAP分析,通过切片、钻取和旋转等方法可从多角度、多侧面去观察数据,从而挖掘出数据中所隐含的规律和知识。更多还原

【Abstract】 There is a large numbers of DCS (Distributed Control System) system data collected in real-time database of thermal power plant. There are abundant and valuable information, which benefits to enhancing the efficiency and safety of operation in thermal power plant, hidden behind these data. But the data resource are wasted generally because not be used for comprehend and mining effectively. The Data Mining techniques can intelligently and automatically discover knowledge from data. In this paper, the application of data mining techniques in thermal power plant will be researched, in the interest of full use the function of DCS system data collected in database to direct the power generation.This paper discussed the applied actuality of data mining techniques in the thermal power plant, and mostly researchs the data preprocessing techniques、the fuzzy c-means algorithm、OLAP(online analytical processing) and the application of these techniques.Considering the characteristic of the data in thermal power plant, this paper attempt to apply the data preprocessing techniques to the data process of real-time database for guarantees the validity of data, so based the foundation for the father work of data mining. This paper attempt to establish the multi-dimensional data model based on the analysis of the condenser performance, and the fuzzy c-means algorithm to be used for the process of the multi-dimensional data. On the base of the data that have been processed, made an multi-dimensional analysis on the vacuum data with the OLAP (online analytical processing) method in the pivot table of Excel, and the data can be viewed in different ways by use the OLAP operations such as slice、roll-up、drill-down and pivot, the rule and the knowledge behind these data can be mined finally.更多还原