Parameter Estimation of Dynamic Beer Fermentation Models

Area of Science:

• Biotechnology
• Chemical Engineering
• Food Science

Background:

• Beer fermentation is a complex biochemical process crucial for product quality.
• Accurate dynamic models are needed for effective process monitoring and control.
• Existing models may lack the predictive power for advanced control strategies.

Purpose of the Study:

• To develop and validate two dynamic models for beer fermentation.
• To estimate model parameters using comprehensive experimental data.
• To assess the models' predictive capabilities for monitoring and control applications.

Main Methods:

• Development of two distinct dynamic mathematical models for beer fermentation.
• Parameter estimation using off-line (biomass, sugar, ethanol, vicinal diketones) and on-line (temperature, ethanol, CO2 exhaust) measurements.
• Structural identifiability analysis to ensure model parameter uniqueness.
• Cross-validation to evaluate model predictive performance.

Main Results:

• Successful estimation of model parameters using diverse experimental data.
• Demonstrated structural identifiability of the proposed models.
• Validated predictive capability of the dynamic models through cross-validation.
• Identified potential applications in advanced process control.

Conclusions:

• The proposed dynamic models offer a robust framework for understanding beer fermentation.
• These models can be utilized as predictors in receding-horizon observers and controllers.
• The study opens new perspectives for real-time monitoring and enhanced control of fermentation processes.