This study explores the application of Machine Learning (ML) models in conjunction with Ultraviolet-Visible spectroscopy for real-time prediction of chemical oxygen demand in industrial wastewater. To this purpose we propose a novel soft sensor architecture that makes use of a supervised regressor based on ensemble of Convolutional Neural Network (CNN) for absorbance spectra and Multi-Layer Perceptron (MLP) for extracted and non-optical features. In our evaluation based on a real dataset built on purpose, beyond models already experimented on these tasks in the literature, we also compare against Recurrent Neural Networks and Echo State Networks that are tailored to process series, such as absorbance spectra. The experimental results show an improvement given by the proposed Hybrid CNN-MLP ensemble model with respect to other models used in the literature.

Hybrid CNN-MLP for Wastewater Quality Estimation

Cardia M.
;
Chessa S.;Micheli A.;
2024-01-01

Abstract

This study explores the application of Machine Learning (ML) models in conjunction with Ultraviolet-Visible spectroscopy for real-time prediction of chemical oxygen demand in industrial wastewater. To this purpose we propose a novel soft sensor architecture that makes use of a supervised regressor based on ensemble of Convolutional Neural Network (CNN) for absorbance spectra and Multi-Layer Perceptron (MLP) for extracted and non-optical features. In our evaluation based on a real dataset built on purpose, beyond models already experimented on these tasks in the literature, we also compare against Recurrent Neural Networks and Echo State Networks that are tailored to process series, such as absorbance spectra. The experimental results show an improvement given by the proposed Hybrid CNN-MLP ensemble model with respect to other models used in the literature.
2024
9783031723551
9783031723568
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11568/1268628
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact