Optimization of the remediation of oil-drilling cuttings by combining artificial neural networks with knowledge-based models

Abstract In the present study, a feed-forward dense multilayer artificial neural network (ANN) trained with backpropagation algorithm (BP) is used to determine kinetic parameters governing performance of oil-drilling cuttings (ODC) ozonation. Ozonation tests ODC, pre-treated surfactant (SDS) and diluted synthetic seawater, are conducted on semi-batch bubble flow reactor. The ozonation experiments reactors evaluated by measuring removal efficiency total organic carbon (TOC). experimental datasets employed calibrate two mathematical models increasing complexity: tank-in-series (TSM), computational fluid dynamics (CFM) model, both combining multiphase transport reactive processes, involving high number unknown parameters, being able provide numerous simulated for various values dimensionless parameters. These data training validation network. then interpreted through Shapley additive explanation (SHAP) method insights about