Advanced search
1 file | 10.21 MB Add to list

Deep learning optimization for soft sensing of hard-to-measure wastewater key variables

(2022) ACS ES&T ENGINEERING. 2(7). p.1341-1355
Author
Organization
Abstract
Soft sensors can be an essential part of a digital twin to acquire critical wastewater information for operation optimization. Soft sensor predictions have been successfully applied in nitrogen compounds, but hard-to-measure variables such as biochemical oxygen demand (BOD) and total suspended solids (TSS) have been a major challenge partially due to difficulty in capturing complex nonlinearity and needed information acquisition. This study pinpointed the bottlenecks by developing advanced hyperparameter optimized (HPO) deep learning (DL) models and testing different groups of data. By comparing two DL algorithms [multilayer perceptron and deep belief network (DBN)] with three HPO methods (genetic algorithm, particle swarm optimization (PSO), and grey wolf optimization), we found that DBN-PSO showed performance superior to other hybrid methods for both CBOD5 and TSS predictions based on 11 years of operational data. While the hybrid models exhibit complex topography, better results can be achieved with a slow learning process and a combination of aggressive pre-training and smooth fine-tuning for CBOD5 and TSS, respectively. Additional precipitation data did not provide additional benefits, whereas metal concentration data helped further improve the prediction accuracy (testing error index: 1.9 mg CBOD5/L and 1.5 mg TSS/L), suggesting that more diverse data acquisition is valuable for a better soft-sensor practice.
Keywords
TREATMENT-PLANT, SUSPENDED-SOLIDS, PREDICTION, NETWORKS, soft sensor, biochemical oxygen demand, total suspended solids, deep, learning, hyperparameter optimization

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 10.21 MB

Citation

Please use this url to cite or link to this publication:

MLA
Zhu, Jun-Jie, et al. “Deep Learning Optimization for Soft Sensing of Hard-to-Measure Wastewater Key Variables.” ACS ES&T ENGINEERING, vol. 2, no. 7, 2022, pp. 1341–55, doi:10.1021/acsestengg.1c00469.
APA
Zhu, J.-J., Borzooei, S., Sun, J., & Ren, Z. J. (2022). Deep learning optimization for soft sensing of hard-to-measure wastewater key variables. ACS ES&T ENGINEERING, 2(7), 1341–1355. https://doi.org/10.1021/acsestengg.1c00469
Chicago author-date
Zhu, Jun-Jie, Sina Borzooei, Jiachun Sun, and Zhiyong Jason Ren. 2022. “Deep Learning Optimization for Soft Sensing of Hard-to-Measure Wastewater Key Variables.” ACS ES&T ENGINEERING 2 (7): 1341–55. https://doi.org/10.1021/acsestengg.1c00469.
Chicago author-date (all authors)
Zhu, Jun-Jie, Sina Borzooei, Jiachun Sun, and Zhiyong Jason Ren. 2022. “Deep Learning Optimization for Soft Sensing of Hard-to-Measure Wastewater Key Variables.” ACS ES&T ENGINEERING 2 (7): 1341–1355. doi:10.1021/acsestengg.1c00469.
Vancouver
1.
Zhu J-J, Borzooei S, Sun J, Ren ZJ. Deep learning optimization for soft sensing of hard-to-measure wastewater key variables. ACS ES&T ENGINEERING. 2022;2(7):1341–55.
IEEE
[1]
J.-J. Zhu, S. Borzooei, J. Sun, and Z. J. Ren, “Deep learning optimization for soft sensing of hard-to-measure wastewater key variables,” ACS ES&T ENGINEERING, vol. 2, no. 7, pp. 1341–1355, 2022.
@article{8769721,
  abstract     = {{Soft sensors can be an essential part of a digital twin to acquire critical wastewater information for operation optimization. Soft sensor predictions have been successfully applied in nitrogen compounds, but hard-to-measure variables such as biochemical oxygen demand (BOD) and total suspended solids (TSS) have been a major challenge partially due to difficulty in capturing complex nonlinearity and needed information acquisition. This study pinpointed the bottlenecks by developing advanced hyperparameter optimized (HPO) deep learning (DL) models and testing different groups of data. By comparing two DL algorithms [multilayer perceptron and deep belief network (DBN)] with three HPO methods (genetic algorithm, particle swarm optimization (PSO), and grey wolf optimization), we found that DBN-PSO showed performance superior to other hybrid methods for both CBOD5 and TSS predictions based on 11 years of operational data. While the hybrid models exhibit complex topography, better results can be achieved with a slow learning process and a combination of aggressive pre-training and smooth fine-tuning for CBOD5 and TSS, respectively. Additional precipitation data did not provide additional benefits, whereas metal concentration data helped further improve the prediction accuracy (testing error index: 1.9 mg CBOD5/L and 1.5 mg TSS/L), suggesting that more diverse data acquisition is valuable for a better soft-sensor practice.}},
  author       = {{Zhu, Jun-Jie and Borzooei, Sina and Sun, Jiachun and Ren, Zhiyong Jason}},
  issn         = {{2690-0645}},
  journal      = {{ACS ES&T ENGINEERING}},
  keywords     = {{TREATMENT-PLANT,SUSPENDED-SOLIDS,PREDICTION,NETWORKS,soft sensor,biochemical oxygen demand,total suspended solids,deep,learning,hyperparameter optimization}},
  language     = {{eng}},
  number       = {{7}},
  pages        = {{1341--1355}},
  title        = {{Deep learning optimization for soft sensing of hard-to-measure wastewater key variables}},
  url          = {{http://doi.org/10.1021/acsestengg.1c00469}},
  volume       = {{2}},
  year         = {{2022}},
}

Altmetric
View in Altmetric