Fault detection and monitoring using Multiscale PCA

In this paper, enhanced method of Principal Component Analysis (PCA) which is Multiscale PCA over conventional PCA to detect fault and monitoring is discussed. Both methods are applied to benchmark ASM 1 Wastewater Treatment Plant (WWTP) system. Then faulty data are detected with SPE and T
²
statistic. The results show the advantages of applying MSPCA over conventional PCA.