Discrimination of Chemical Oxygen Demand Pollution in Surface Water Based on Visible Near-Infrared Spectroscopy

Chemical oxygen demand (COD) is one of the indicators used to monitor level pollution in surface water. To recycle agricultural water resources, it crucial monitor, a timely manner, whether COD exceeds control standard. A diagnostic model was developed using visible near-infrared spectroscopy (Vis-NIR) combined with partial least squares discriminant analysis (PLS–DA). total 127 samples were collected from Guangzhou, Guangdong, China. The content measured potassium dichromate method. spectra recorded Vis-NIR spectrometer, and spectral data pre-processed four different methods. improve accuracy simplicity model, synthetic minority oversampling technique (SMOTE) competitive adaptive reweighted sampling (CARS) algorithm enhance performance. best PLS–DA achieved an 88%, SMOTE–PLS–DA had 94%. SMOTE could despite imbalance. CARS–SMOTE–PLS–DA 97% accuracy, CARS band selection improved discrimination model. by 9% over that This method can not only save human material resources but also new way for real-time online