Machine Learning Methods for the Prediction of Wastewater Treatment Efficiency and Anomaly Classification with Lack of Historical Data

This study examines an algorithm for collecting and analyzing data from wastewater treatment facilities, aimed at addressing regression tasks for predicting the quality of treated wastewater and classification tasks for preventing emergency situations, specifically filamentous bulking of activated s...

Full description

Saved in:
Bibliographic Details
Published in:Applied sciences 2024-11, Vol.14 (22), p.10689
Main Authors: Gulshin, Igor, Kuzina, Olga
Format: Article
Language:English
Subjects:
Algorithms
Data collection
effluent quality
Energy consumption
Machine learning
machine learning algorithms
Methods
Neural networks
Nitrogen
Purification
Sensors
Sewage
Sludge
soft sensors
Time series
wastewater treatment
Water treatment
Water treatment plants
Water utilities
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study examines an algorithm for collecting and analyzing data from wastewater treatment facilities, aimed at addressing regression tasks for predicting the quality of treated wastewater and classification tasks for preventing emergency situations, specifically filamentous bulking of activated sludge. The feasibility of using data obtained under laboratory conditions and simulating the technological process as a training dataset is explored. A small dataset collected from actual wastewater treatment plants is considered as the test dataset. For both regression and classification tasks, the best results were achieved using gradient-boosting models from the CatBoost family, yielding metrics of SMAPE = 9.1 and ROC-AUC = 1.0. A set of the most important predictors for modeling was selected for each of the target features.
ISSN:2076-3417
2076-3417
DOI:10.3390/app142210689