Software sensors for biomass concentration in a SSC process using artificial neural networks and support vector machine

• Published in: Bioprocess and biosystems engineering 2014-01, Vol.37 (1), p.27-36
• Main Authors: Acuña, Gonzalo, Ramirez, Cristian, Curilem, Millaray
• Format: Article
• Language: English
• Subjects: Algorithms; Biomass; Biotechnology; Carbon dioxide; Carbon Dioxide - chemistry; Cell Culture Techniques - methods; Chemistry; Chemistry and Materials Science; Computer Simulation; Environmental Engineering/Biotechnology; Fermentation; Food Science; Fungi - metabolism; Gibberellins - chemistry; Industrial and Production Engineering; Industrial Chemistry/Chemical Engineering; Neural networks; Neural Networks (Computer); Nonlinear Dynamics; Original Paper; Oxygen - chemistry; Regression Analysis; Reproducibility of Results; Sensors; Software; Support Vector Machine; Vector space
• ISSN: 1615-7591; 1615-7605
• DOI: 10.1007/s00449-013-0925-3

The lack of sensors for some relevant state variables in fermentation processes can be coped by developing appropriate software sensors. In this work, NARX-ANN, NARMAX-ANN, NARX-SVM and NARMAX-SVM models are compared when acting as software sensors of biomass concentration for a solid substrate cultivation (SSC) process. Results show that NARMAX-SVM outperforms the other models with an SMAPE index under 9 for a 20 % amplitude noise. In addition, NARMAX models perform better than NARX models under the same noise conditions because of their better predictive capabilities as they include prediction errors as inputs. In the case of perturbation of initial conditions of the autoregressive variable, NARX models exhibited better convergence capabilities. This work also confirms that a difficult to measure variable, like biomass concentration, can be estimated on-line from easy to measure variables like CO 2 and O 2 using an adequate software sensor based on computational intelligence techniques.