A Gas Sensors Detection System for Real-Time Monitoring of Changes in Volatile Organic Compounds during Oolong Tea...

Area of Science:

• Food Science and Technology
• Analytical Chemistry
• Sensor Technology

Background:

• Oolong tea quality is critically dependent on the oxidation process, influencing its characteristic flavor and aroma.
• Monitoring oxidation stages is essential for consistent Oolong tea production.
• Volatile organic compounds (VOCs) are key indicators of the oxidation status.

Purpose of the Study:

• To develop and validate a gas sensor detection system for real-time monitoring of Oolong tea oxidation.
• To analyze the volatile aroma compounds released during oxidation.
• To establish effective machine learning models for classifying oxidation stages.

Main Methods:

• Utilized a gas sensor system with 13 metal oxide semiconductor sensors.
• Collected and analyzed volatile organic compound (VOC) signals during oxidation.
• Employed Principal Component Analysis (PCA) for feature extraction.
• Applied Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Back-Propagation Neural Network (BP-ANN), LeNet5, and AlexNet for signal recognition.

Main Results:

• The sensor system effectively detected VOCs like aldehydes, alcohols, and olefins released during oxidation.
• The Back-Propagation Neural Network (BP-ANN) model demonstrated superior performance.
• BP-ANN achieved high accuracy rates of 94.16% (calibration) and 94.11% (prediction) for oxidation stage recognition.
• The system successfully differentiated between distinct stages of the Oolong tea oxidation process.

Conclusions:

• The proposed gas sensor detection system provides an effective method for monitoring Oolong tea oxidation.
• This technology can significantly improve Oolong tea product stability and consistency.
• The system facilitates the automation of the oxidation process through long-term, online, real-time monitoring.