Prediction of a full scale WWTP activated sludge SVI test using an LSTM neural network
- 17 September 2022
- journal article
- research article
- Published by Royal Society of Chemistry (RSC) in Environmental Science: Water Research & Technology
- Vol. 8 (11) , 2786-2795
Abstract
The two key parameters to control and prevent sludge separation problems in activated sludge are SVI test and microscopic observations. To predict SVI we propose a bidirectional LSTM model which shows an accuracy rate of 89% and an F1-score of 82%.Keywords
Funding Information
- Microsoft (AI for earth grant)
This publication has 19 references indexed in Scilit:
- Image Processing and Analysis of Phase-Contrast Microscopic Images of Activated Sludge to Monitor the Wastewater Treatment PlantsIEEE Access, 2017
- Prediction of sludge volume index bulking using image analysis and neural network at a full-scale activated sludge plantDesalination and Water Treatment, 2015
- A fuzzy neural network approach for online fault detection in waste water treatment processComputers and Electrical Engineering, 2014
- Automatic identification of activated sludge disturbances and assessment of operational parametersChemosphere, 2013
- Microscopic Image Analysis versus Sludge Volume Index to Monitor Activated Sludge Bioflocculation: A Case StudySeparation Science and Technology, 2013
- Prediction of activated sludge bulking based on a self-organizing RBF neural networkJournal of Process Control, 2012
- Dynamic modeling of filamentous bulking in lab-scale activated sludge processesJournal of Process Control, 2006
- Manual on the Causes and Control of Activated Sludge Bulking, Foaming, and Other Solids Separation Problems2003
- Long Short-Term MemoryNeural Computation, 1997
- Relationships between biological solids retention time and settling characteristics of activated sludgeWater Research, 1971