A baseline correction model for humidity and temperature compensationWO3 film based sensor for NO2 detection

In this work, a generalized self-adaptive algorithm implementation to remove the humidity dependent gas sensor response of metal-oxide based sensors has been investigated. The sensor output analysis has been divided into steps of baseline stabilization, maximum deviation search, leak detection and recovery. The baseline search using consecutive slope comparison is accurate with an error of 0.225 to 26.88 %. Using an adaptive baseline model, the saturation value for 3 ppm NO
₂
is accurately detected with a maximum error of 2.7 % and response magnitudes were found to have negligible variation. Using Ordinary Least Squares and QR Factorization, the baseline in the presence of gas was predicted using historical data of temperature and humidity relationship with current to find the point of recovery of baseline, hence considerably reducing the humidity & temperature effect which can be applied to all metal oxide-based gas sensors.