2021
DOI: 10.1016/j.watres.2021.117697
|View full text |Cite
|
Sign up to set email alerts
|

Prediction of biogas production in anaerobic co-digestion of organic wastes using deep learning models

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

1
30
1

Year Published

2022
2022
2024
2024

Publication Types

Select...
4
2

Relationship

0
6

Authors

Journals

citations
Cited by 55 publications
(32 citation statements)
references
References 29 publications
1
30
1
Order By: Relevance
“…The detailed information on the sampling frequency and distribution of the data can be found in the supporting information. However, the key parameters that represent the composition of AD feedstock, such as BOD, COD, and TN (Jeong et al, 2021; Karamichailidou et al, 2022; Wang et al, 2021), were not measured at the SIUE WWTP. In this study, if a parameter was measured periodically, the measured value was used to represent the entire duration.…”
Section: Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…The detailed information on the sampling frequency and distribution of the data can be found in the supporting information. However, the key parameters that represent the composition of AD feedstock, such as BOD, COD, and TN (Jeong et al, 2021; Karamichailidou et al, 2022; Wang et al, 2021), were not measured at the SIUE WWTP. In this study, if a parameter was measured periodically, the measured value was used to represent the entire duration.…”
Section: Methodsmentioning
confidence: 99%
“…There are very few works to predict biogas production at full-scale WWTPs using ML algorithms such as Random Forest (RF) models, eXtreme Gradient Boosting (XGBoost), and K-Nearest Neighbor (KNN) regression (de Clercq et al, 2019(de Clercq et al, , 2020Wang et al, 2021). A success story in this regard is the development of a deep learning model that predicted biogas production from anaerobic codigestion processes in a full-scale WWTP based on 17 process features collected over 2 years, including biological oxygen demand (BOD), chemical oxygen demand (COD), total nitrogen (TN), and total phosphorus (TP) in the primary and secondary sludges as well as the sludge feeding into AD operator (Jeong et al, 2021). Such a success is built on the quality set of data, which as mentioned earlier may not be the case in many small WWTPs.…”
Section: Introductionmentioning
confidence: 99%
“…While the linear regression models have simple and understandable inputs, the co-digestion process likely contained nonlinearities that were not captured by simple models. ML models have the drawback that they provide no insight into mechanistic connections between independent variables and observations; however, they are more robust to co-correlated variables, such as those found in time-series problems such as this one and have been frequently applied for biogas prediction. ,, …”
Section: Methodsmentioning
confidence: 99%
“…ML models have the drawback that they provide no insight into mechanistic connections between independent variables and observations; however, they are more robust to co-correlated variables, such as those found in time-series problems such as this one and have been frequently applied for biogas prediction. 7,10,13 We used an automated ML (AutoML) tool to create a robust ML pipeline without biases for model selection, feature standardization, and hyperparameter optimization. To do so, we used the Tree-based Pipeline Optimization Tool (tpot library in Python), which selects feature preprocessing operators, conducts feature selection, combines optimal models, and conducts hyperparameter optimization using genetic programming to compile an ML pipeline.…”
Section: Tree-based Pipeline Optimizationmentioning
confidence: 99%
See 1 more Smart Citation