Development of a real-time expert system for wastewater treatment plants control

Abstract

A real-time expert system to control wastewater treatment plants is presented. The software has been developed in the G2 environment. It contains: an interface that permits on-line acquisition of plant data using G2 standard interface (GSI), a predictive control algorithm for dissolved oxygen (DO) control, and a graphical interface between the expert system and the operator. The dissolved oxygen control is performed using a non-linear predictive control algorithm, that has been developed to satisfy quality constraints whilst reducing energy demands. The algorithm uses data obtained from the plant by hardware sensors, and software which recursively estimates the oxygen uptake rate. All these elements are integrated in a knowledge base that includes a set of diagnosis, detection, prediction and operation rules, making the system capable of handling a wide number of usual (where predictive control can be useful) and unusual situations (where quantitative and qualitative information must be considered.