摘要
在介绍过程监控主要内容及方法的基础上,归纳分析了统计过程监控的发展,详细阐述了传统的PCA、PCR、PLS、FDA、HMM和几种改进的PCA方法.指出,将人的经验和专家知识融入到PCA统计过程监控,提高PCA的故障检测、隔离能力,完善故障诊断与评价能力,将是一个有价值的研究方向.
The main contents and methods of process monitoring are introduced. The development of statistical process monitoring is analyzed. The traditional PCA, PCR, PLS, FDA, HMM and some improved PCA methods are elaborated in details. Some conclusions and further research works are given. A valuable research direction is to combine the human experience and expert knowledge with PCA statistical process monitoring to raise the PCA ability of fault detection and isolation, improving the capacity of fault diagnosis and evaluation.
基金
国家自然科学基金资助项目(60274027)
关键词
统计过程监控
PCA
数据驱动
Statistical process monitoring
principal component analysis
data-driven