摘要
提出了化工生产过程数据挖掘系统的技术架构,在此基础上,采用多支持度关联规则挖掘算法,开发了化工生产过程数据挖掘软件ESP-Miner,该软件提供OPC Server标准接口,主流实时数据库接口、关系数据库接口ODBC、标准文本格式接口,可集成DCS、实时数据库系统、关系数据库系统及其他系统的生产数据,同时提供人工录入接口集成散点数据。该系统已经在三唑磷合成过程中得到了成功的应用,三唑磷收率提高了1.5%,三唑磷含量从80%提高到了83%,为企业带来了较好的经济效益和社会效益。
The technical frame of data-mining system for chemical process is proposed. Based on algorithm of association rules with multiple minimum supports, the data-mining software ESP-Miner for chemical process is developed. The software can integrate process data from Distributed Control System（DCS） , real-time database system, relational database system and other system by standard interface OPC Server, interface for mainstream real-time database, ODBC interface for relational database and standard text format interface. Furthermore, process data can be input by man-machine conversion interface. The ESP-Miner software is successfully carried out in triazophos synthesis process. The production indicates that after data mining recipe and parameter are adopted, the product yield of triazophos is increased by 1.5% on average, and the average triazophos content increases from 80% to 83%. It brings about more economic benefits and social benefits for corporation.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2008年第7期769-772,共4页
Computers and Applied Chemistry
基金
国家创新研究群体科学基金(NCRGSFC:60721062)
国家高技术研究发展计划资助项目(863计划2007AA04Z172).
关键词
数据挖掘
关联规则
实时数据库
三唑磷合成过程
data-mining, association rules, real-time database, triazophos synthesis process
