On 1 October 2024, Mr Arne Freyschmidt successfully defended his dissertation on the topic of "N₂O emission reduction through AI-supported control of biological wastewater treatment" and received the highest possible rating of "with distinction".
Mr Freyschmidt's central scientific approach is to replace classic methods of process simulation with neural networks or to couple them in a hybrid manner in order to realise further automated and forecast-based process plant control with the aim of incorporating new additional control variables (emission minimisation). In his dissertation, Mr Freyschmidt explicitly investigated the approach of using an optimisation algorithm based on a neural network. The genetic algorithm used is intended to identify the control setting for specific operating situations that results in minimum N2O emissions. Mr Freyschmidt's work thus demonstrates a new way of achieving optimised control for the operation of biological wastewater treatment by combining deterministic models and neural networks with the support of AI.
The doctorate was completed as part of the research projects "Minimising the CO2 footprint through adapted process development in process water treatment - testing the MiNzE process in a submerged fixed bed (MiNzE)" and "Process technologies in the main and satellite operation of an inter-municipal recycling centre for optimised regional nutrient recycling (SATELLITE)", which were funded by the BMBF.
We would like to congratulate Arne Freyschmidt on his doctorate!
