Temperature and Humidity Compensation for MOS Gas Sensor Based on Random Forests

The outputs of Metal Oxide Semiconductor (MOS) gas sensors drift due to the change of temperature and humidity in the environment. This phenomenon leads to additional errors in the measurement and the test precision and measurement stability of gas sensor are greatly affected. A novel strategy for temperature and humidity compensation for MOS Gas Sensor is proposed in this paper. The environmental gas concentrations are measured separately and accurately based Random Forest (RF) method to demonstrate that the proposed strategy is superior at both accuracy and runtime compared with the conventional methods, such as RBF neural network and BP neural network. Results show that the proposed methodology provides a better solution to temperature and humidity drift. The accuracy of the environmental gas sensor array improves about 1%.