Data-driven modelling of N2O production in wastewater processes using neural ordinary differential equations

Data-driven modelling of extended activated sludge processes with greenhouse gases (ASMG1 model, including N2O production) in a BSM1 plant. Please check the paper below for details: "Data-driven modelling of N2O production in wastewater processes using neural ordinary differential equations".

"/data/" folder under root contains training data for different influent scenarios (dry, rain, storm)

run "main_training.m" file under "/matrix_version_v5_metrics/" folder for training, it may take a few days depending on the size of training data you choose.

run "main_testing.m" file under "/matrix_version_v5_metrics/" folder for testing a trained model. The parameters of the trained model is saved in the "matrix_version_v5_metrics/data/" folder. There is an example, just try it.