TY - JOUR
T1 - Modeling phosphorous dynamics in a wastewater treatment process using Bayesian optimized LSTM
AU - Hansen, Laura Debel
AU - Stokholm-Bjerregaard, Mikkel
AU - Durdevic, Petar
PY - 2022/4
Y1 - 2022/4
N2 - This study presents a systematic framework to develop data-driven models for phosphorus concentration in a full-scale wastewater treatment plant (WWTP). The dynamics of wastewater treatment exhibit nonlinear behavior, and are time varying, non-stationary, and coupled in a complex manner, which makes them difficult to predict using mechanistic models. Two long short-term memory (LSTM) models are proposed. The first estimates the phosphorus concentration using data describing environmental conditions and process operation, and the second model which additionally utilizes the previous phosphorus measurement. Additionally, the hyperparameters are tuned using Bayesian optimization, as this is an effective tool to determine the best model and prevent over-fitting and long training duration of the data-driven models. The two models show good prediction performances and are suitable to predict up to 24 hours into the future, with R
2 close to 0.7-0.8 for data well presented in the training data set.
AB - This study presents a systematic framework to develop data-driven models for phosphorus concentration in a full-scale wastewater treatment plant (WWTP). The dynamics of wastewater treatment exhibit nonlinear behavior, and are time varying, non-stationary, and coupled in a complex manner, which makes them difficult to predict using mechanistic models. Two long short-term memory (LSTM) models are proposed. The first estimates the phosphorus concentration using data describing environmental conditions and process operation, and the second model which additionally utilizes the previous phosphorus measurement. Additionally, the hyperparameters are tuned using Bayesian optimization, as this is an effective tool to determine the best model and prevent over-fitting and long training duration of the data-driven models. The two models show good prediction performances and are suitable to predict up to 24 hours into the future, with R
2 close to 0.7-0.8 for data well presented in the training data set.
KW - Control
KW - Deep Learing
KW - LSTM
KW - Machine Learning
KW - Model
KW - Nitrogen
KW - Waste Water Treatment
KW - Dynamic model
KW - Full scale plant data
KW - Hyperparameter tuning
KW - Neural networks
KW - Phosphorus
KW - Time series prediction
UR - http://www.scopus.com/inward/record.url?scp=85125119808&partnerID=8YFLogxK
U2 - 10.1016/j.compchemeng.2022.107738
DO - 10.1016/j.compchemeng.2022.107738
M3 - Journal article
SN - 0098-1354
VL - 160
JO - Computers & Chemical Engineering
JF - Computers & Chemical Engineering
M1 - 107738
ER -