Signal model of electronic noses with metal oxide semiconductor

• We developed a new signal model for electronic nose with metal oxide semiconductor.
• The model relates the signal intensity and the absorbed odorant on sensors.
• The mathematical expression of the model is not a bilinear or trilinear form.

Electronic noses have found wide applications as a promising tool in many fields. However, a model for their signals has not been clearly elucidated thus far. Here, we have developed a signal model for electronic noses based on the relationship between the signal intensity and the amount of odorant molecules absorbed onto the surface of a sensor. We applied an alternative least square algorithm to solve the model. We successfully applied this model to the analysis of benzene and six perfumes using data from a commercial system.