Artificial neural network real‐time process control system for small utilities

Abstract

This article presents the final results from an Awwa Research Foundation project to implement an artificial intelligence real‐time automatic control system for a full‐scale dissolved‐air flotation (DAF) water treatment plant. The system developed through study used an artificial neural network to model the DAF clarifier and was integrated with the plant supervisory control and data acquisition system for real‐time control. The system was found to be robust and flexible, and it adapted well to fast‐changing raw water conditions and different coagulants. Once established, the system ran the plant continuously, as if an expert were in control, providing quality drinking water with minimum operator intervention. The system was also capable of optimizing chemical dosage and reducing the costs of production.

Real Time Prediction of WWTP Inflow Using Cascade Correlations and Genetic Algorithm ANN: Developing an Early Warning System for High Flows at Gold Bar WWTP

“…al. 2007 andStanley 1999). In Alberta the software has been used successfully in modelling stream flow and water quality parameters in natural streams, optimizing chemical water treatment processes, and other water infrastructure processes.…”

“…Sample Model Verification Results for Modelling Alternative #5 ANN Alternative #5 Performance(Validation Data from 2003, 2005, and 2007 …”