Application of black-box models based on artificial intelligence for the prediction of chlorine and TTHMs in the trunk network of Bogotá, Colombia

Abstract The chlorine and total trihalomethane (TTHM) concentrations are sparsely measured in the trunk network of Bogotá, Colombia, which leads to a high uncertainty level at an operational level. For this reason, research assessed prediction accuracy for TTHM two black-box models based on following artificial intelligence techniques: neural networks (ANNs) adaptive neuro-fuzzy inference system (ANFIS) as modelling alternative. simulation results hydraulic water quality analysis EPANET its multi-species extension EPANET-MSX were used training models. Subsequently, Threat Ensemble Vulnerability Assessment-Sensor Placement Optimization Tool (TEVA-SPOT) Evolutionary Polynomial Regression-Multi-Objective Genetic Algorithm (EPR-MOGA-XL) jointly applied select most representative input variables locations predicting other points network. ANNs ANFIS optimized with multi-objective approach reach compromise between performance generalization capacity. had higher mean Training Test Nash–Sutcliffe Index (NSI) contrast ANNs. In general, satisfactory performance. However, some them did not achieve suitable NSI values, different statuses was limited.