In this paper we present the application of adaptive predictive expert controllers to dissolved oxygen (DO) control in the aerobic reactors of a wastewater treatment plant. The control system described in this paper consists of adaptive predictive expert control loops complemented by optimisation logic. The controllers successfully cope with nonlinearity and changing operating conditions of the process by predicting the evolution of the controlled variable and adapting to changes in the process dynamics. This results in more precise and stable DO control, offering many benefits. The complementary optimisation logic maintains the air pressure in the common collector at the lowest possible level, enabling adequate DO control and thus considerably reducing energy consumption.
Keywords: activated sludge process, adaptive predictive control, dissolved oxygen control, energy consumption reduction