DETECTION OF UNMEASURED PROCESS ABNORMALITIES IN WASTEWATER TREATMENT PROCESS USING MS-PCA

MultiScale – Principal Component Analysis (MS-PCA) is a very suitable method for fault detection of wastewater treatment processes. It will be shown that MS-PCA can notice non-measurable process abnormalities significantly faster than a human eye. Simulation results based on bacteria population decrease case will be presented.