Search

This paper presents the application of a robust non-diagonal multivariable control design method, based on the Quantitative Feedback Theory (QFT), to the control of a complex wastewater treatment plant (WWTP) with biological removal of nitrogen and phosphorus. The control objective is to simultaneously regulate the concentration of ammonia, nitrates and phosphorus in the plant effluent. The manipulated variables are the dissolved oxygen (DO) level in the aerobic reactor, the sludge recycling flow from the settler and the amount of ferric hydroxide added to chemically precipitate the phosphorus. Thus, the system becomes a three-input-three-output control structure with model parameter uncertainty. A robust non-diagonal multivariable QFT method is used to design a fully populated matrix controller. This sequential method reduces the interactions between control loops, increases the robust stability and fulfils the robust performance specifications defined by the user. It also considers non-minimum phase aspects by avoiding the introduction of right-half plane transmission zeros due to the controller elements. Simulations carried out by using the IWA ASM2d model under WEST® software validate the proposed control methodology.