论文部分内容阅读
由于基于主元分析(Principal Component Analysis,PCA)的统计监控方法没有利用过程机理模型(First Principle Model)信息,因此在一定程度上限制了其故障诊断能力的发展。本文基于PCA的框架,采用故障子空间对故障进行描述,在PCA监测模型的基础之上,分析了主元空间和残差空间的故障可检测性问题,获得了故障可检测性的必要充分理论条件。通过对双效蒸发过程的仿真监测,证实了所获理论结果的有效性,表明了通过计算临界故障幅值就可事先对故障集内各故障的检测结果作定量的分析,从而事先了解各故障在PCA下的检测结果。
Because the statistical monitoring method based on Principal Component Analysis (PCA) does not use First Principle Model information, it limits the development of its fault diagnosis ability to a certain extent. Based on the framework of PCA, the fault subspace is used to describe the fault. Based on the PCA monitoring model, the fault detectability of the principal component space and residual space is analyzed. The necessary and sufficient theory of fault detectability is obtained. condition. Through the simulation monitoring of the double-effect evaporation process, the validity of the theoretical results obtained is confirmed, which shows that the quantitative analysis of the detection results of each fault in the fault set can be obtained in advance by calculating the critical fault amplitude so as to know each fault in advance Test results under PCA.