Department Process Engineering

Project: EMISSUN

Efficient Model Identification for Soft-Sensing of Urine Nitrification


A novel integrated modelling framework is proposed for modelling of biological wastewater treatment processes. The modelling framework is generally applicable to any biological process and combines (i) extent-based modelling as developed and studied at the Automation Laboratory at EPFL (Prof. Dominique Bonvin), (ii) shape constrained splines as flexible model building blocks studied earlier for qualitative trend analysis with the Spike group at Eawag (Dr. Kris Villez), and (iii) state-of-the-art parameter estimation techniques, selected and fine-tuned following consultation with the SIAM department at Eawag.


Partners:Automation Control Laboratory (EPFL)
Spike staff:Alma Mašić