PCA在过程故障检测与诊断中的应用

讨论了基于主无分析（PCA）的过程故障检测与诊断的原理，运用T^2统计、Q统计方法，结合贡献图对一典型过程进行了仿真分析，结果表明PCA方法可对简单传感器故障进行检测与诊断，并指出了该方法中的不足，提出了将PCA方法同基于过程动态模型的故障诊断方法相结合的研究思路。

Fault Detection and Diagnosis Based on Principal Component Analysis

Fault detection and diagnosis method based on the principal components analysis (PCA) is discussed . The fault detection and diagnosis simulation to a typical chemical process is performed by means of statistical methods like Holleting T 2 and Q. The contribution charts are used to undertake fault diagnosis. The simulation results show that PCA is an effective approach to fault detection and can only work for the simple sensor fault diagnosis. A novel idea to combine PCA with causal model based approach is presented for the future research aiming at complex sensor fault and internal process fault.