[21a-B01-1~10] 23.1 Joint Session N "Informatics"

[21a-B01-5] Development of a Gas Sensor System Based on Transfer Function Ratios Toward Artificial Olfaction

〇Gaku Imamura¹, Kota Shiba¹, Genki Yoshikawa^1,2, Takashi Washio³ (1.NIMS, 2.Univ. of Tsukuba, 3.Osaka Univ.)

Keywords:machine olfaction, signal analysis, Membrane-type Surface stress Sensor (MSS)

In this study, we developed a new data processing method based on transfer function ratios (TFRs), which is independent of gas flow control. To demonstrate the feasibility of this method, we performed gas sensing measurements by using Membrane-type Surface stress Sensors (MSS). Using TFRs as features, we successfully developed a gas classification model that is applicable to measurement data obtained with different gas flow control. Furthermore, we demonstrated facile odor identification based on TFRs through “free-hand measurement”, in which odor is identified simply by holding a gas sensor chip near a sample.