Temperature Effect on Capacitive Humidity Sensors and its Compensation Using Artificial Neural Networks
Tarikul Islam, Zaheer Uddin, Amit Gangopadhyay · 2015
This paper represents the study of the effect of temperature on different capacitive humidity sensors used in practice. Capacitance of the humidity sensor, which is a function of concentration of water vapor, also depends on ambient temperature. This variation of ambient temperature causes error in the performance of sensor outputs and its compensation is essential. In this paper, we have used an artificial neural network to compensate the effect of ambient temperature error. The proposed artificial neural network technique is based on inverse model of the sensor. The technique is applicable for compensation of linear or nonlinear temperature effect of humidity sensor. It can also compensate the nonlinearity of the capacitive humidity response which is an issue for all most all types of humidity sensor. Our simulation studies show the sensor output and artificial neural network model output matches closely. Even though sensor characteristics change with temperature, the proposed model performs well irrespective of any change in temperature. It can be extended for the temperature compensation of other sensors. The maximum error for nonlinearity using the ANN technique are 0.2 % and temperature error of 0.08 % for temperature range between 10 °C to 60 °C of Sensor 3 and 0.01 % for temperature range between 25 °C to 85 °C of Sensor 4 respectively. Copyright © 2015 IFSA Publishing, S. L.