RESUMEN

Fourier transform near-infrared (FT-NIR) spectroscopy combined with multivariate analysis has been applied in bioprocesses for a couple of decades. Nevertheless the papers published in this field are case-specific and do not focus on providing the community generic workflows to conduct experiments, especially as a standard Design of Experiment (DoE) for a multi-analyte process might require overwhelming amount of measurements. In this paper, a workflow for feasibility studies and inline implementation of FT-NIR spectrometer in multi-analyte fermentation processes is presented. The workflow is applied to Penicillium crysogenum fermentation, where the similarities in chemical structures and growth trends between the key analytes together with the aeration and growing fungi make the task challenging: first, the pure analytes are measured off-line with FT-NIR and clustered using principal component analysis. To study the separability of the gained clusters, a DoE approach by spiking is applied. The multivariate modelling of the separable analytes is conducted using the off-line and inline data followed by a comparison of the properties of the different models. Finally, the model output constraints are set by means of outlier diagnostics. As a result, biomass, penicillin (PEN), phenoxyacetic acid (POX), ammonia and biomass were shown to be separable with root mean square error of predictions of 2.62 g/l, 0.34 g/l, 0.51 g/l and 18.3 mM, respectively. Graphical abstract Flowchart illustrating the workflow for feasibility studies and implementation of models for inline monitoring of Ammonia, Biomass, Phenoxyacetic acid and Penicillin.

Asunto(s)

RESUMEN

In viscose production, it is important to monitor three process parameters in order to assure a high quality of the final product: the concentrations of H(2)SO(4), Na(2)SO(4) and Z(n)SO(4). During on-line production these process parameters usually show a quite high dynamics depending on the fiber type that is produced. Thus, conventional chemometric models, which are trained based on collected calibration spectra from Fourier transform near infrared (FT-NIR) measurements and kept fixed during the whole life-time of the on-line process, show a quite imprecise and unreliable behavior when predicting the concentrations of new on-line data. In this paper, we are demonstrating evolving chemometric models which are able to adapt automatically to varying process dynamics by updating their inner structures and parameters in a single-pass incremental manner. These models exploit the Takagi-Sugeno fuzzy model architecture, being able to model flexibly different degrees of non-linearities implicitly contained in the mapping between near infrared spectra (NIR) and reference values. Updating the inner structures is achieved by moving the position of already existing local regions and by evolving (increasing non-linearity) or merging (decreasing non-linearity) new local linear predictors on demand, which are guided by distance-based and similarity criteria. Gradual forgetting mechanisms may be integrated in order to out-date older learned relations and to account for more flexibility of the models. The results show that our approach is able to overcome the huge prediction errors produced by various state-of-the-art chemometric models. It achieves a high correlation between observed and predicted target values in the range of [0.95,0.98] over a 3 months period while keeping the relative error below the reference error value of 3%. In contrast, the off-line techniques achieved correlations below 0.5, ten times higher error rates and the more deteriorate, the more time passes by.