Artificial neural networks application for analysis and control of grapes fermentation process

Wine's high economic value requires the development of methods to assess its quality. Anyway, winemaking is often managed by operators that decode the physicochemical phenomena occurring during manufacture. Integrating their skills in a control framework would represent an enhancement of the quality control procedure. The anthocyanin amount in wine is influenced by the fermentation process parameters, is responsible for red wine colour and suitable for intensity and tonality classification. In this paper, an artificial neural network was implemented to predict the intensity and tonality of wine's colour as a function of different parameters, which occur during the fermentation process.