On neuro-fuzzy applications for automatic control, supervision, and fault diagnosis for water treatment plant

Abstract

Water treatment includes many complex phenomena, such as coagulation and flocculation. These reactions are hard or even impossible to control satisfyingly by conventional methods. Biological water treatment systems are difficult to model because their performance is complex and varies significantly with different reactor configurations, influent characteristics, and operational conditions. Neuro-fuzzy ANFIS method, which is chosen as the method in this case, is a new intelligent method in this line of process industry. Although intelligent tools such as neural network, fuzzy logic and neuro-fuzzy methods have been applied in real time water treatment plant for some time, problems of monitoring water treatment processes and assessing uncertainty for the coagulant dosing rate represent a major challenged that need to be investigated. In this research, statistical methods are used to analyze nonstationary time series water treatment process where they are accrued from a neuro-fuzzy ANFIS model. The proposed scheme is evaluated in computer simulation studies using real process data before application to the real plant.