Vygandas Vaitkus
ABSTRACT
Efficient control of bioprocess is becoming more and more relevant in today's biotechnology industry and environment of changing technologies. Many of important bioprocess variables are not measured on-line. Usually it took 15 min and some of them only after 24 hours (Dry biomass). On-line estimation of unknown bioprocess parameters provide improved process control performance. In this article, three advanced techniques for soft-sensors design were investigated: support vector regression, relevance vector regression and random forest regression model. As direct measurements for estimation were used glucose/lactose feed rates and oxygen uptake rate along with its integrated quantity. Estimation quality of models analyzed in first stage was tested by using data of mechanistic process model. Later models were tested on E. coli BL21 (DE3) pET21-IFN-alfa-5 strain cultivation process measurements. As inputs were used: Time after induction, CPR (Carbon dioxide Production Rate) and Feeding flow along with its integrated values. Outputs of the models were: OD (Optical Density) and acetate quantity. All models provide results close to each other but random forest showed slightly better efficiency.
- B Sonnleitner, 1999 Instrumentation of biotechnological processes in Advances in Biochemical Engineering and Biotechnology, Springer, 3--54.Google Scholar
- CF Mandenius, 2004 Recent developments in the monitoring, modeling and control of biological production systems Bioproc. Biosyst. Eng vol 26 pp 347--351.Google Scholar
- R Simutis, V Galvanauskas, D Levisauskas, J Repsyte and V Vaitkus, 2013 Comparative Study of Intelligent Soft-Sensors for Bioprocess State Estimation. JOLST, 163--167.Google Scholar
- P Kadlec, B Gabrys and S Strandt, 2009 Data-driven soft sensors in the process industry Comput. Chem. Eng vol 33, 795--814.Google ScholarCross Ref
- MM Zhang and XG Liu, 2013 A soft sensor based on adaptive fuzzy neural network and support vector regression for industrial melt index prediction, Chemometrics and Intelligent Laboratory Systems, vol. 126, 83--90.Google ScholarCross Ref
- R Mansano, E Godoy, A Porto, 2014 The Benefits of Soft Sensor and Multi-Rate Control for the Implementation of Wireless Networked Control Systems, Sensors, 14, 24441--24461.Google ScholarCross Ref
- GC Goodwin, 2000 Predicting the performance of soft sensors as a route to low cost automation, Annual Reviews in Control, 24, 55--66.Google ScholarCross Ref
- M Nrgaard and PM Norgaard, 2006 Neural Networks for Modelling and Control of Dynamic Systems, A Practitioner's Handbook (Advanced Textbooks in Control and Signal Processing), Springer.Google Scholar
- V Galvanauskas, R Simutis, N Volk, A Lübbert, 1998 Model based design of a biochemical cultivation process, Bioprocess and Biosystems Engineering, vol 18, no 3, 227--234.Google ScholarCross Ref
- V Galvanauskas, N Volk, R Simutis, A Lübbert, 2004 Design of recombinant protein production processes, Chemical Engineering Communications, vol. 191, no 5, 732--748.Google ScholarCross Ref
- CC Chang and CJ Lin, 2001 LIBSVM: A Library for Support Vector Machines.Google Scholar
- V Vapnik, 1998 Statistical Learning Theory, Wiley, New York, NY.Google Scholar
- B Schölkopf, AJ Smola, 2002 Learning with Kernels, MIT Press Cambridge MA.Google Scholar
- ME Tipping, 2001 Sparse bayesian learning and the relevance vector machine, Journal of Machine Learning Research, vol. 1, 211--244.Google ScholarDigital Library
- C Mills Terence, 1990 Times Series Techniques for Economists, Cambridge University Press.Google Scholar
- M Belgiu and L Dragu, 2016 Random forest in remote sensing: A review of applications and future directions, ISPRS Journal of Photogrammetry and Remote Sensing, 24--31.Google Scholar
- S Bernard, L Heutte and S Adam, 2009 Influence of hyperparameters on random forest accuracy, In MCS vol 5519 of Lecture Notes in Computer Science, Springer, 171--180.Google ScholarCross Ref
Index Terms
- Soft-sensors based on Black-box Models for Bioreactors Monitoring and State Estimation
Recommendations
Aged Leachate Treatment by Sequencing Aged Refuse Bioreactors
ESIAT '09: Proceedings of the 2009 International Conference on Environmental Science and Information Application Technology - Volume 03In this study, the aged refuse excavated from a 10-year old closed landfill compartment at Shanghai Laogang Refuse Landfill was used as the substrate in a sequencing aged refuse bioreactor process for treatment of aged landfill leachate which sampled ...
Procedures for parameter and state estimation of microbial growth process models
Practical problems concerning parameter and state estimation of microbial growth processes, with application to batch and continuous fermentation processes and to the activated sludge wastewater treatment process are discussed. Besides the usual ...
Using machine learning techniques to predict ammonium concentration in membrane contactors for nitrogen recovery as a valuable resource
AbstractThe availability of real-time measurements of primary quality water variables is one of the key challenges in the wastewater treatment industry. However, due to the cost and maintenance requirements of sensors and probes for on-line measurement ...
Graphical abstractDisplay Omitted
Highlights- Real-time pH measurements were used to accurately predict the real-time evolution of TAN concentration.
- ANN outperformed the other machine learning approaches, exhibiting RMSE <20 mg/L.
- The relationship between the target and the ...
Comments