Abstract
This paper describes the study of a dissolved oxygen (DO) concentration control in an activated sludge process. The DO concentration at a specific point is controlled around the time-varying set point, which is so manipulated that the gain of the DO concentration with respect to the aeration air flow rate becomes maximum. A new discrete-time algorithm is developed for this control. This algorithm is based on high-pass filters characterized by the parallel configuration of a time lag and a moving average, a difference-form broken line fitting for the gain maximization, and an adaptive PI control, with which the DO concentration is controlled around the time-varying set point. Experiments in a pilot-scale plant show that this control algorithm performs satisfactorily to be applied in actual plants.