A novel gas sensor signal drift adjustment method based on controlled measurement

摘要

Gas sensors of semiconductor types are more or less affected by external environmental factors change, which could inevitably lead to signal drift of measurements. In this study, we propose a novel method to adjust sensor signal drift. The system design includes: (1) real-time feedback of remote environmental factors by wireless sensor networks; (2) a controlled measurement setup used for simulating changeable remote environmental factors; (3) drift adjustment of remote sensor signal based on controlled sensor signal. We process extracted sensitivity feature vector of three cases with Na?ve Bayes classifier for comparison, and acquired experimental results demonstrate that our drift adjustment method can effectively improve recognition rate at the presence of sensor signal drift. Especially, this design is appropriate for remote control sensor applications under changeable environmental factors.