Gross Solids Content Prediction in Urban WWTPs Using SVM

Abstract

The preliminary treatment of wastewater at wastewater treatment plants (WWTPs) is of great importance for the performance and durability of these plants. One fraction that is removed at this initial stage is commonly called gross solids and can cause various operational, downstream performance, or maintenance problems. To avoid this, data from more than two operation years of the Villapérez Wastewater Treatment Plant, located in the northeast of the city of Oviedo (Asturias, Spain), were collected and used to develop a model that predicts the gross solids content that reaches the plant. The support vector machine (SVM) method was used for modelling. The achieved model precision (Radj2 = 0.7 and MSE = 0.43) allows early detection of trend changes in the arrival of gross solids and will improve plant operations by avoiding blockages and overflows. The results obtained indicate that it is possible to predict trend changes in gross solids content as a function of the selected input variables. This will prevent the plant from suffering possible operational problems or discharges of untreated wastewater as actions could be taken, such as starting up more pretreatment lines or emptying the containers.

Using a supervised machine learning approach to predict water quality at the Gaza wastewater treatment plant

“…26 SVM-based prediction has been extensively researched for tracking and forecasting intake conditions and sludge volume index in wastewater treatment plants. 27 An adaptive multi-output so sensor model, 28 hybrid linear-nonlinear method, 29 data-based predictive control technique, 30 and SVM model 31,32 have also been used to predict effluent index, total solid content, and water quality of wastewater treatment facilities. The least-squares support vector machine (LSSVM) has been presented as a solution to the drawbacks of SVM for large datasets.…”