Stormwater runoff is often contaminated by human activities. Stormwater discharge into\nwater bodies significantly contributes to environmental pollution. The choice of suitable treatment\ntechnologies is dependent on the pollutant concentrations. Wastewater quality indicators such as\nbiochemical oxygen demand (BOD5), chemical oxygen demand (COD), total suspended solids (TSS),\nand total dissolved solids (TDS) give a measure of the main pollutants. The aim of this study is\nto provide an indirect methodology for the estimation of the main wastewater quality indicators,\nbased on some characteristics of the drainage basin. The catchment is seen as a black box: the\nphysical processes of accumulation, washing, and transport of pollutants are not mathematically\ndescribed.Two models deriving from studies on artificial intelligence have been used in this research:\nSupport Vector Regression (SVR) and Regression Trees (RT). Both the models showed robustness,\nreliability, and high generalization capability. However, with reference to coefficient of determination\nR2 and root-mean square error, Support Vector Regression showed a better performance than\nRegression Tree in predicting TSS, TDS, and COD. As regards BOD5, the two models showed a\ncomparable performance. Therefore, the considered machine learning algorithms may be useful for\nproviding an estimation of the values to be considered for the sizing of the treatment units in absence\nof direct measures.
Loading....