Ethanol Concentration in Yeast Batch Cultivation via Gas Sensor Array

A theoretical model was used to simulate ethanol production in Saccharomyces cerevisiae batch cultivation for sensor calibration. The root mean square error of calibration and percentage error for validation sets were below 0.2 g/L and 7%, respectively. The dataset was created by Abdolrahim Yousefi Darani at the University of Hohenheim.

Use Cases

• Calibrating gas sensor arrays for on-line ethanol prediction based on voltage changes.
• Validating theoretical process models for yeast cultivation without off-line sampling.
• Comparing sensor-based parameter acquisition to traditional least squares fitting methods.

Strengths

• Model-based calibration achieved a root mean square error below 0.2 g/L.
• Validation error was below 7%, indicating good predictive performance.
• Calibration procedure required no off-line sampling, enabling continuous monitoring.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Hohenheim

Collection Method

Head space samples were pumped past a gas sensor array every five minutes for 10 seconds, and voltage changes were measured.