Neural network based fault detection and identification of a neutralisation process control system

Abstract

This dissertation presents a Neural Network Based Fault Detection and Identification (FDI) Scheme for an existing neutralisation process control system in Sembcorp Utilities and Terminals Ltd, Singapore. The neutralisation process plant is to convert the alkaline water of PH factor around 9 into neutralised water of PH factor around 7. The neutralisation process is controlled by a Programmable Logical Controller, using block and bleed arrangement of three numbers of ON/OFF valves.