Intelligent tetrahydrothiophene gas detection based on electrochemical sensor array.

Area of Science:

• Chemical Engineering
• Analytical Chemistry
• Sensor Technology

Background:

• Natural gas odorization is crucial for detecting leaks and preventing explosions.
• Electrochemical sensors face cross-interference challenges, impacting accuracy in odorant detection.
• Accurate identification and concentration measurement of odorants are vital for gas safety.

Purpose of the Study:

• To develop a reliable method for natural gas odorant identification and concentration detection.
• To overcome the limitations of conventional methods and sensor cross-interference.
• To ensure early warning systems for natural gas without compromising gas quality.

Main Methods:

• Utilized Principal Component Analysis (PCA) and K-nearest neighbor (KNN) algorithm for gas recognition.
• Employed a backpropagation-AdaBoost model with an electrochemical sensor array for tetrahydrothiophene concentration estimation.
• Validated the method using natural gas samples from Chengdu's Chenghua district.

Main Results:

• Achieved a gas recognition rate of 90.17%.
• Reduced the average relative error in tetrahydrothiophene concentration detection to 3.37%.
• Demonstrated significant improvement in prediction accuracy and mitigation of cross-interference effects.

Conclusions:

• The proposed method effectively enhances the accuracy of natural gas odorant detection.
• It provides a reliable technical solution for identifying and quantifying tetrahydrothiophene, addressing gas safety concerns.
• The approach offers significant engineering value for improving natural gas safety protocols.