Using historical data for bioprocess optimization: modeling wine characteristics using artificial neural networks and archived process information.

Optimization of fermentation processes is a difficult task that relies on an understanding of the complex effects of processing inputs on productivity and quality outputs. Because of the complexity of these biological systems, traditional optimization methods utilizing mathematical models and statistically designed experiments are less effective, especially on a production scale. At the same time, information is being collected on a regular basis during the course of normal manufacturing and process development that is rarely fully utilized. We are developing an optimization method in which historical process data is used to train an artificial neural network for correlation of processing inputs and outputs. Subsequently, an optimization routine is used in conjunction with the trained neural network to find optimal processing conditions given the desired product characteristics and any constraints on inputs. Wine processing is being used as a case study for this work. Using data from wine produced in our pilot winery over the past 3 years, we have demonstrated that trained neural networks can be used successfully to predict the yeast-fermentation kinetics, as well as chemical and sensory properties of the finished wine, based solely on the properties of the grapes and the intended processing. To accomplish this, a hybrid neural network training method, Stop Training with Validation (STV), has been developed to find the most desirable neural network architecture and training level. As industrial historical data will not be evenly spaced over the entire possible search space, we have also investigated the ability of the trained neural networks to interpolate and extrapolate with data not used during training. Because a company will utilize its own existing process data for this method, the result of this work will be a general fermentation optimization method that can be applied to fermentation processes to improve quality and productivity.

[1]  P. Weimer,et al.  Response surface analysis of the effects of pH and dilution rate on Ruminococcus flavefaciens FD-1 in cellulose-fed continuous culture , 1992, Applied and environmental microbiology.

[2]  B. K. Lonsane,et al.  Computerized study of interactions among factors and their optimization through response surface methodology for the production of tannin acyl hydrolase by Aspergillus niger PKL 104 under solid state fermentation , 1994 .

[3]  Jules Thibault,et al.  Comparison of prediction performances between models obtained by the group method of data handling and neural networks for the alcoholic fermentation rate in enology , 1991 .

[4]  A. Bull,et al.  Vancomycin production in batch and continuous culture , 2000, Biotechnology and bioengineering.

[5]  J. Macgregor,et al.  Development of inferential process models using PLS , 1994 .

[6]  Jean-Marie Sablayrolles,et al.  Alcoholic fermentation under oenological conditions , 1995 .

[7]  M. N. Karim,et al.  Comparison of feed-forward and recurrent neural networks for bioprocess state estimation , 1992 .

[8]  J Glassey,et al.  Artificial neural network based experimental design procedure for enhancing fermentation development , 1994, Biotechnology and bioengineering.

[9]  B. C. Bhattacharyya,et al.  Evolutionary operation (EVOP) to optimize three‐dimensional biological experiments , 1993, Biotechnology and bioengineering.

[10]  Satoshi Imayasu,et al.  Analysis of the state characteristics of sake brewing with a neural network , 1992 .

[11]  Antonio M. López,et al.  A simple mathematical empirical model for estimating the rate of heat generation during fermentation in white-wine making , 1992 .

[12]  Jer-Yiing Houng,et al.  Search method for the optimal medium for the production of lactase by Kluyveromyces fragilis , 1992 .

[13]  Tohru Suzuki,et al.  Production of xylitol from D‐xylose by candida tropicalis: Optimization of production rate , 1992, Biotechnology and bioengineering.

[14]  D. Cantero,et al.  Development of a kinetic model for the alcoholic fermentation of must , 1991, Biotechnology and bioengineering.

[15]  W. Fred Ramirez,et al.  Optimization of Fed‐Batch Bioreactors Using Neural Network Parameter Function Models , 1996 .

[16]  G. Montague,et al.  Enhanced supervision of recombinant E. coli fermentation via artificial neural networks , 1994 .

[17]  G. T. Tsao,et al.  Neural network modeling of batch cell growth pattern , 1993, Biotechnology and bioengineering.

[18]  Sheyla L. Rivera,et al.  Global and local neural network models in biotechnology: Application to different cultivation processes , 1997 .

[19]  Bhavik R. Bakshi,et al.  Representation of process trends—IV. Induction of real-time patterns from operating data for diagnosis and supervisory control , 1994 .

[20]  W. Fred Ramirez,et al.  A FLAVOUR MODEL FOR BEER FERMENTATION , 1994 .

[21]  Pascal Formisyn,et al.  Malolactic fermentation of wine: study of the influence of some physico-chemical factors by experimental design assays , 1995 .

[22]  Hung-Chang Chen,et al.  Optimizing the concentrations of carbon, nitrogen and phosphorus in a citric acid fermentation with response surface method , 1996 .

[23]  W. Fred Ramirez,et al.  Neural‐network modeling and optimization of induced foreign protein production , 1999 .

[24]  B. Schenker,et al.  Cross-validated structure selection for neural networks , 1996 .

[25]  A. J. Morris,et al.  Bioprocess model building using artificial neural networks , 1991 .

[26]  Cornelius S. Ough,et al.  Methods for analysis of musts and wines , 1980 .

[27]  A. J. Morris,et al.  Towards improved penicillin fermentation via artificial neural networks , 1992 .

[28]  S. Sánchez,et al.  Optimization of a culture medium for streptomycin production using response-surface methodology , 1993 .

[29]  J Thibault,et al.  On‐line prediction of fermentation variables using neural networks , 1990, Biotechnology and bioengineering.

[30]  X M Yang,et al.  Optimization of a cultivation process for recombinant protein production by Escherichia coli. , 1992, Journal of biotechnology.

[31]  G Stephanopoulos,et al.  Fermentation database mining by pattern recognition. , 1997, Biotechnology and bioengineering.

[32]  Taizo Hanai,et al.  Temperature control of Ginjo sake mashing process by automatic fuzzy modeling using fuzzy neural networks , 1998 .

[33]  Q. Chen,et al.  Optimization for a recombinantE. coli fed-batch fermentation , 1995, Applied biochemistry and biotechnology.