The invention provides an electronic
nose online analysis method in gas-sensitive-
gas chromatography multi-
perception information selection, fusion and
fermentation / malodor processes. The electronic
nose online analysis method is characterized in that: a gas sample injection single period T<0> is 300-600s, the detected gas sample injection time of a gas sensitive
sensor array module is asynchronous from the detected gas sample injection time of a capillary gas
chromatographic column module, the accumulated sample injection amount of the gas sensitive
sensor array module is unequal to the accumulated sample injection amount of the capillary gas
chromatographic column module, and the sensing information selection and the analysis time are synchronous; 3 sensing information, namely a steady-state
voltage peak value, a peak-appearing time and an
area under the curve, is selected by a
computer control and analysis module from a single gas sensitive sensor time length 60s response curve, and21 sensing information, namely 10 maximum peak values, 10 retention times and 1
area under the curve, is selected from a time length T<0>-10s semi-separation chromatogram; a modular deep convolutionneural network is utilized by the electronic
nose instrument to realize long-term online detection and analysis of 5
fermentation tanks or malodor
pollution points with the maximum cycle period of T=5T<0>, including
odor type identification, overall
odor intensity and main component concentration quantitative prediction.