Stochastic Predictive Control of Wastewater Treatment Processes

Peter Stentoft: Wastewater treatment plants are extremely important for modern societies. They protect local ecosystems from large nutrient loads and reduce human health risks related to exposure to faecal matter. However the operation of a wastewater treatment plant is very resource demanding in terms of electricity, chemicals, maintenance etc. and thus control is essential.

There exist several rule-based control (RBC) strategies which have proven good at maintaining a steady, sufficient performance of the plant. However, the RBC does not automatically manage external changes. Consequently if one wishes to keep an optimal operation of the WWTP, the RBC needs to be updated when changes in the loads, maintenance, energy price, legislation etc. occur. Since energy prices and loads (and possibly other effects) vary on a daily basis, the job of updating the RBC manually becomes impractical.

The goal of this project is to develop stochastic model predictive control strategies to control the wastewater treatment plants. These data-driven strategies will always adapt to the current state of the plant, and it is therefore expected that they will outperform current control. Furthermore, the predictive approach introduces the possibility of using energy prices as an input and prioritize green energy. It is therefore expected that this project will lead to new control software which will ultimately reduce operational costs on wastewater treatment plants.

PhD project title: Stochastic Predictive Control of Wastewater Treatment Processes

Effective start/end date 15/09/2017 → 14/09/2020

DTU supervisors: Jan Kloppenborg Møller and Henrik Madsen for the section for Dynamical Systems.

 

PhD project by Peter Stentoft

Research section: Dynamical Systems

Principal supervisor: Jan Kloppenborg Møller

Co-supervisor: Henrik Madsen, Peter Steen Mikkelsen, Thomas Munk-Nielsen & Luca Vezzaro

Title of project: Stochastic Predictive Control of Wastewater Treatment Processes

Project start: 15/09/2017 → 14/09/2020

Contact

Jan Kloppenborg Møller
Associate Professor
DTU Compute
+45 45 25 34 18

Contact

Henrik Madsen
Professor, Head of section
DTU Compute
+45 45 25 34 08