Abstract: Principal component analysis (PCA) has found wide application in chemical process monitoring and product quality control as a data-driven modeling method.Based on the fault reconstruction technology, the fault isolation and identification issues were explored by using T² index, since some faults involving process fault and sensor fault that do not violate the dominant process relation described by the PCA statistical model could only be detected by T² index.The theoretical conditions of isolatability and identifiability were obtained.The acquired results were then illustrated and verified by monitoring a simulated double-effect evaporator (DEE) process, where 10 different faults were studied.

Key words: 主元分析, 产品质量控制, 故障分离, 故障识别

摘要： 主元分析 (PCA)作为数据驱动的一种统计建模方法，在化工产品质量控制与故障诊断方面得到广泛研究和应用.在故障重构技术的基础上，研究了基于T²统计量的故障分离和识别问题，分别获得了主元空间中故障可分离和识别的理论条件.以双效蒸发过程为例，对该生产过程中的10种不同故障进行仿真监测分析，证实了所获理论结果的有效性.

关键词: 主元分析, 产品质量控制, 故障分离, 故障识别