•  
  •  
 

Abstract

This paper presents the development and investigation of hybrid neural network and fuzzy models for the mathematical modelling of nitrification and denitrification processes in a biological wastewater treatment bioreactor. A comprehensive approach is proposed, integrating a mechanistic model of the ASM1/ASM2d type with neural networks (LSTM and Gaussian Process Regression), as well as a fuzzy control system based on an extended set of expert rules. A digital twin of the bioreactor was developed to allow for the prediction of the dynamic behavior of key parameters such as NH₄⁺, NO₃⁻, dissolved oxygen, etc. within a prediction range of 1 to 12 hours. The study also includes a stability analysis of the hybrid control system based on Lyapunov criteria, as well as a proof-of-concept evaluation of the system performance under temperature fluctuations, sensor noise, and sudden organic matter events.  The results show that the accuracy of predicting ammonium concentration using neural network models was significantly improved, i.e., the MSE was reduced by 38-55%, while the integration of fuzzy logic enabled automatic adjustment of control actions and reduced aeration energy consumption by 12-18%. The proposed architecture can serve as a foundation for the implementation of intelligent control systems within the framework of Industry 4.0 digital wastewater treatment facilities.

First Page

44

Last Page

49

References

  1. Henze, M., Gujer, W., Mino, T., & Van Loosedrecht, M. (2006). Activated sludge models ASM1, ASM2, ASM2d and ASM3. IWA publishing. https://doi.org/10.2166/9781780402369
  2. Rittmann, B. E., & McCarty, P. L. (2001). Environmental biotechnology: principles and applications (Vol. 6). New York: McGraw-Hill. https://abe.ufl.edu/media/abeufledu/documentsx2fpdf/course-syllabi/ABE4932_EnvironmentalBioetchnologyHWCOESyllabus_Spring2026-Pratap-Pullammanappallil.pdf
  3. Tchobanoglous, G., Burton, F.L. and Stensel, H.D. (2003) Wastewater Engineering, Treatment and Reuse. 4th Edition, McGraw-Hill, Boston. Journal of Water Resource and Protection, Vol.7 No.15, October 20, 2015
  4. Е. Талмурзин, А.Кизатова, G.Bazil, и А.Ербосынова, «СОВРЕМЕННЫЕ ТЕНДЕНЦИИ И ИННОВАЦИИ В АВТОМАТИЗИРОВАННОМ УПРАВЛЕНИИ ПРОЦЕССАМИ БИОЛОГИЧЕСКОЙ ОЧИСТКИ ВОДЫ», Вестник КазАТК, т. 143, вып. 2, апр. 2026.
  5. Jeppsson, U., Pons, M. N., Nopens, I., Alex, J., Copp, J. B., Gernaey, K. V., ... & Vanrolleghem, P. A. (2007). Benchmark simulation model no 2: general protocol and exploratory case studies. Water Science and Technology, 56(8), 67-78. https://doi.org/10.2166/wst.2007.604.  
  6. Makinia, J., & Zaborowska, E. (2020). Mathematical modelling and computer simulation of activated sludge systems. IWA publishing. DOI: https://doi.org/10.2166/9781780409528, February 2020.
  7. Tirez, Arne & Stevens, Niels & Bongartz, Dominik & Assumpcao, José. (2026). Hybrid Modeling of Wastewater Treatment Dynamics Using Hammerstein-Wiener Structures. https://doi.org/10.69997/sct.192067.
  8. Lin, W., Hanyue, Y., & Bin, L. (2022). Prediction of wastewater treatment system based on deep learning. Frontiers in ecology and evolution, 10, 1064555. doi: 10.3389/fevo.2022.1064555  
  9. Ospina Alarcón, M. A., Chanchí Golondrino, G. E., & Úsuga Manco, L. M. (2025). Machine Learning Algorithms for Dynamic System Identification in Wastewater Treatment Plant. http://hdl.handle.net/20.500.14044/31939 .
  10. Alamu, R., Karkala, S., Hossain, S., Krishnapatnam, M., Aggarwal, A., Zahir, Z., ... & Shah, V. (2025). Physics-Informed Neural Networks for Climate Modeling: Bridging Machine Learning and Physical Laws.
  11. Wang, Y. Q., Wang, H. C., Song, Y. P., Zhou, S. Q., Li, Q. N., Liang, B., ... & Wang, A. J. (2023). Machine learning framework for intelligent aeration control in wastewater treatment plants: Automatic feature engineering based on variation sliding layer. Water Research, 246, 120676. doi:10.1016/j.watres.2023.120676.  
  12. Hvala, N., & Kocijan, J. (2020). Design of a hybrid mechanistic/Gaussian process model to predict full-scale wastewater treatment plant effluent. Computers & Chemical Engineering, 140, 106934. DOI:10.1016/j.compchemeng.2020.106934
  13. Janiesch, C., Zschech, P., & Heinrich, K. (2021). Machine learning and deep learning: C. Janiesch et al. Electronic markets, 31(3), 685-695. https://doi.org/10.1007/s12525-021-00475-2
  14. Computation, N. (2016). Long short-term memory. Neural Comput, 9, 1735-1780. Doi: 10.1162/neco.1997.9.8.1735.
  15. Pinto, J., Mestre, M., Ramos, J., Costa, R. S., Striedner, G., & Oliveira, R. (2022). A general deep hybrid model for bioreactor systems: Combining first principles with deep neural networks. Computers & Chemical Engineering, 165, 107952. Doi: 10.1016/j.compchemeng.2022.107952.
  16. Pang, J., Yang, S., He, L., Chen, Y., & Ren, N. (2019). Intelligent control/operational strategies in WWTPs through an integrated Q-learning algorithm with ASM2d-guided reward. Water, 11(5), 927. https://doi.org/10.3390/w11050927
  17. Spielberg, S. P. K., Gopaluni, R. B., & Loewen, P. D. (2017, May). Deep reinforcement learning approaches for process control. In 2017 6th international symposium on advanced control of industrial processes (AdCONIP) (pp. 201-206). IEEE.  DOI: 10.1109/ADCONIP.2017.7983780.
  18. Bozdogan, H. (1987). Model selection and Akaike's information criterion (AIC): The general theory and its analytical extensions. Psychometrika, 52(3), 345-370. https://doi.org/10.1007/BF02294361

Download

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.