Process Fault Detection and Reconstruction by Principal Component Analysis

Qi, R; Zhang, J

doi:10.1109/MMAR.2019.8864614

Process Fault Detection and Reconstruction by Principal Component Analysis

Lookup NU author(s): Dr Ruosen Qi, Dr Jie Zhang ORCiD

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

Existing fault reconstruction methods are very effective in dealing with sensor faults not involved in control loops where the fault direction is usually easy to determine. However, implementing fault reconstruction methods for process faults or sensor faults involved with control loops is quite challenging as the fault direction vectors are usually difficult to specify. Process faults usually affect a number of process variables with various extents. This paper introduces a principal component analysis (PCA) based fault reconstruction method for process faults. PCA is used to analyze historical process data with faults to extract fault directions, which are then used for fault reconstruction. The proposed method is demonstrated on a simulated continuous stirred tank reactor.