摘要
化工生产过程往往含有大量的过程变量,且过程多处于闭环控制作用下,产生的测量数据常常存在互相关和自相关。规范变量分析(CVA)通过最大化两个变量集间的相关度,实现对高维数据的降维,并得到一组最大限度地解释变量集中信息的规范变量,很好地解决了上述问题。本文介绍一种基于CVA的过程监控方法,并将此方法应用于一实际化工单元的过程监控,利用控制图,及时准确地检测到过程故障,表明了基于CVA的监控方法的有效性。
Chemical industrial processes always have a large number of process variables and are usually operated under closed-loop control, so the process data measured often exist crosscorrelation and autocorrelation. Canonical variate analysis （CVA） is a dimensionality reduction technique to solve the above problem perfectly through maximizing the correlation degree between two sets of variables. Meanwhile, the canonical variates are obtained to maximumly explain the information in variables sets. In this paper, a process monitoring method based on CVA is introduced to monitor a practical chemical process. Combined with statistic chart, CVA can achieve the purpose of real time monitor of the whole process. The results show that this approach is an efficient process monitoring method.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2007年第2期247-250,共4页
Computers and Applied Chemistry
基金
新世纪优秀人才支持计划(NCET-05-0485)
