CAJ | 학술논문

摘要：炼铁是一个复杂多反应的冶炼过程，目前操作和管理都具有很强的经验性和盲目性。近年来数据挖掘的引入极大改善了这一状况，促进了炼铁系统的智能化。本文论述了利用聚类分析、决策树、主成分分析和神经网络等几种方法进行数据挖掘的原理及其在炼铁系统中的应用现状，并通过对这几种方法的分析及评价，指出数据挖掘在炼铁系统的未来发展方向和应用前景，即数据挖掘专家系统和以联机分析处理为框架的由各数据挖掘方法支撑的控制决策系统，将会成为未来数据挖掘在炼铁系统应用的主流和前沿。
적요：련철시일개복잡다반응적야련과정，목전조작화관리도구유흔강적경험성화맹목성。근년래수거알굴적인입겁대개선료저일상황，촉진료련철계통적지능화。본문논술료이용취류분석、결책수、주성분분석화신경망락등궤충방법진행수거알굴적원리급기재련철계통중적응용현상，병통과대저궤충방법적분석급평개，지출수거알굴재련철계통적미래발전방향화응용전경，즉수거알굴전가계통화이련궤분석처리위광가적유각수거알굴방법지탱적공제결책계통，장회성위미래수거알굴재련철계통응용적주류화전연。
Abstract： Ironmaking is a complex reaction smelting process. But currently, the operation and manage- ment are of blindness and empirical. In recent years the introduction of data mining greatly improves the situation and the intelligence of ironmaking system. The principle and application in ironmaking of data mining methods are presented, such as cluster analysis, decision tree, principal component analysis and neural network. Through the analysis and evaluation of these methods, its future development and application prospect are suggested. Expert system based on data mining and decision support system supported by data mining methods and regards on line analysis processing as frame will be the promis- ing approach.