Abstract
Wastewater treatment processes with activated sludge are described in the specialized literature by complex models with
nonlinear parameterization, such as for example Activated Sludge Model ASM1, ASM2 or ASM3. Under these conditions, the
design of control structures using the state space representation is very difficult. Suitable techniques to approach the control of
these processes are using control structures based on an input-output model or using control structures obtained without even
knowing the process model. In this paper two techniques of this type are analyzed: a data driven technique, Virtual Reference
Feedback Tuning (VRFT), and a robust control technique, Quantitative Feedback Theory (QFT). The control structures designed
by the two methods are implemented using a wastewater treatment plant implemented in the simulation software SIMBA for
which a complex influent was considered. The influent includes information on water temperature and gives data for a period of
one year. The analysis of the two methods considers the quality of the obtained control results but, at the same time, the difficulty
of implementing the two methods.