A Novel Gas Identification Method Based on Gabor Spectrogram Using Self-adapted Temperature Modulated Gas Sensors

In this paper, a self-adapted temperature modulation circuit, whose period and magnitude of the output voltage depend on the online resistances of the sensor array, along with a feature extraction method based on Gabor spectrogram is proposed to improve the identification ability of the MOX sensors. The Gabor spectrogram of the sensor response shows that the direct current (DC) component is heavily dependent on the steady-state response of the gas sensor associated with the concentration and the type of target gas, while the high frequency components have a close relationship with the frequency of the operating temperature. Therefore, features of the sensor responses are extracted from the Gabor spectrogram at the DC frequency to fulfill the identification task. The proposed algorithm is evaluated on the analysis of six kinds of volatile drinks. Experimental results show that the proposed method is not only effective in the signal analysis, but also useful and applicable for the performance enhancement of the current electronic noses.