Identification of process operating state with operational map in municipal wastewater treatment plant.

This work was performed to develop an operational map for the objective diagnosis of the process operating states of a municipal wastewater treatment plant, for which multivariate statistical analysis techniques were applied. PCA (principal component analysis) was used to reduce the dimension of the data sets obtained from the field municipal wastewater treatment plant. A K-means clustering analysis was used to classify the group according to the property of the process operating state. A Fisher's linear discriminant analysis was used to derive the discriminant function of each classified group. An operational map was developed by scatter-plotting the derived principal components (PCs) on a two-dimensional coordinate according to the classified groups. Using the new data sets not used for developing the operational map, the practical usefulness of the operational map and discriminant function in diagnosing the process operating state were evaluated. Hence, the process operating state could be easily and quickly diagnosed and the dynamic trend of the process operating state was also able to be estimated using the operational map.