Correction Model for Metal Oxide Sensor Drift Caused by Ambient Temperature and Humidity

Area of Science:

• Materials Science
• Sensor Technology
• Environmental Monitoring

Background:

• Metal oxide (MOX) gas sensors offer high sensitivity, broad detection range, fast response, and cost-effectiveness, making them suitable for smart city applications, gas monitoring, and safety systems.
• However, the performance of MOX gas sensors can be significantly impacted by environmental factors such as ambient temperature and humidity, leading to cross-sensitivity and unreliable readings.
• Understanding and mitigating these environmental effects is crucial for enhancing the accuracy and dependability of MOX gas sensor applications.

Purpose of the Study:

• To investigate the cross-sensitivity of different MOX gas sensors (MiCS-5524, GM-402B, GM-502B, and MiCS-6814) to ambient temperature and humidity variations.
• To develop and validate regression models capable of correcting the sensor response drift caused by these environmental factors.
• To demonstrate the effectiveness of the proposed models in improving the stability and reliability of MOX gas sensor outputs.

Main Methods:

• A gas sensor array was constructed, incorporating temperature and humidity sensors alongside four distinct MOX gas sensors.
• The sensor array was exposed to varying concentrations of gases under controlled temperature (16 °C to 30 °C) and humidity (75% to 45%) conditions, simulating a typical indoor environment.
• Regression models were developed for each MOX sensor to mathematically correlate and compensate for the observed drifts in sensor response due to temperature and humidity fluctuations.

Main Results:

• Gas sensor responses were found to be significantly influenced by ambient temperature and humidity. Increased temperature and humidity generally led to decreased sensor response, with the MiCS-6814 sensor exhibiting an inverse relationship.
• The developed regression models successfully corrected the sensor response drift, resulting in substantially reduced standard deviations compared to the raw sensor outputs.
• Validated models demonstrated a significant minimization of drift, with corrected standard deviations being considerably lower than the raw sensor response values (e.g., 1.66 kΩ vs. 18.22 kΩ for one sensor).

Conclusions:

• Ambient temperature and humidity significantly affect the performance of MOX gas sensors, necessitating compensation strategies for accurate measurements.
• The proposed regression models effectively mitigate the drift in MOX gas sensor responses caused by temperature and humidity variations, leading to more stable and reliable outputs.
• The developed correction models are applicable to the specific MOX sensors tested and show potential for broader use with other MOX gas sensors in various applications, including training processes.