【作者】 包洋；

【导师】 李蓓智；

【作者基本信息】 东华大学 ， 机械制造及自动化， 2006， 硕士

【摘要】 数据挖掘是致力于数据分析和理解、揭示数据内部蕴藏知识的技术,它成为未来信息技术应用的重要目标之一。经过十几年的努力,数据挖掘产生了许多新概念和方法。特别是最近几年,一些基本概念和方法趋于清晰,它的研究正向着更深入的方向发展。像其它新技术的发展历程一样,数据挖掘技术也必须经过概念提出、概念接受、广泛研究和探索、逐步应用和大量应用等阶段。同时,现代制造业随着工艺、装备和信息技术的不断发展而发展,也产生了大量的有价值的数据,如何高效的利用这些数据为企业服务便日益成为人们关注的焦点。 本文在对数据挖掘技术以及企业生产过程中的数据特点进行充分分析的基础上,提出了在制造业生产领域中充分利用数据挖掘技术,为企业生产加工中的工艺参数优化,过程检测等方面提供解决方案的构想,并研制一套数据挖掘系统工具平台来对此构想进行具体的实现。本文具体完成的工作主要有以下几个方面: 首先,通过对数据挖掘技术在实施过程、系统组成等多个方面的研究与分析,揭示出数据挖掘与传统的数据分析工具之间的差异,使对数据挖掘的理解以及其目的意义更加的明晰。 其次,在解决工艺参数优化的问题中,本文提出了一种正演的方法,即结合使用决策树分类器以及人工神经网络进行综合分析的方法来完成。同时,这种方法还为类似由指标要求反求多个因素的优化组合问题给出了一个很好的解决思路。 再次,在数据挖掘算法的研究直至实现过程中,根据制造业数据的特点,对关联挖掘算法进行改进,使其对多维的制造业数据也能行之有效的进行分析,从而拓展了关联算法应用对象的范围。 最后,在数据挖掘系统工具平台研究方面,按照面向对象的思想更多还原

【Abstract】 Data mining is a technology that aims to analyze and understand large source data and reveal knowledge hidden in the data. It has been viewed as an important evolution in information processing. During the past decade or over, the concepts and technology on data mining have been presented, and some of them have been discussed in higher levels for the last few years. And it becomes the focus of attention that how do use these data to provide service to enterprises efficiently.The data mining in the production process technology and enterprise data characteristics on the basis of full analysis, presented in the field to take full advantage of manufacturing production data mining techniques, the production and processing enterprises optimize processes parameters, such as the testing process for the solution concept and the development of a data mining system platform specific ideas to achieve this. The concrete work completed the following main areas:Firstly, through data mining technology implement course , system person who make up a lot of research and analysis of, reveal the data mining with the difference between the traditional data analysis tools, make to understanding and purpose and meaning their that data mining distinct.Secondly, in solving the problem that the craft parameter is optimized, this paper has put forward a method to perform , which using decision tree and ANN carry on comprehensive. Meanwhile, this kind ofmethod , still in order to demand to ask the optimization grouping problems of a lot of factors to provide a very good thinking of settlement similarly instead by the index.Thirdly, while the research of data mining algorithm until realizing, according to the characteristic of the data of manufacturing industry, to improve the associate algorithm , make it enable effectual analyzing, thus expanded the range of employing the target of associate algorithm.Finally, the research of the data mining system tool, carry on the overall design to the scheme of the whole tool platform according to the target OOP, set up the systematic system structure and function model of the data mining system, building up for the systematic platform of the form of expression with C/S structure, realized the main key algorithm of the module of the data mining system.更多还原