Advancements in machine learning for estimating parameters of wastewater treatment plants

The aim of the study is to develop and validate machine learning methods for calculating the parameters of aeration tanks of wastewater treatment plants at the stage of technical and commercial proposal. Research methods included: generalization of known scientific and technical results, theoretical studies were conducted using the theory of fluid motion in the boundary layer, the theory of kinetics of enzymatic reactions of organic pollutants in wastewater, machine learning methods and statistical decision theory. Experimental studies were conducted on a laboratory setup to study the kinetics of wastewater sedimentation. As a result of the study, a model of the XGBoost algorithm was developed, which successfully coped with the task of optimization of calculations, providing high accuracy, and this, in turn, opens up new opportunities for improving the efficiency of design of wastewater treatment plants.