Imaged liquor recognition method applicable to multiple electronic nose platforms

Abstract

The invention relates to an imaged liquor recognition method applicable to multiple electronic nose platforms. The method comprises the steps that liquor sample original data collected by different sensors is acquired and subjected to normalization processing; sensor response values are converted into an angle cosine and matrix through an anti-cosine function; a half-angle sine and matrix Hs and ahalf-angle cosine and matrix Hc are obtained; Markov transfer probabilities are calculated after binning operation is performed on sensor responses, and then a Markov domain matrix MTF is calculated;sensor images are constructed, piecewise aggregate approximation is utilized to determine image dimensions, and multiple sensor images are combined for each sample to obtain a combination image describing the single sample; a convolutional neural network model is established; and training is performed.

Description

technical field

[0001] The invention relates to an imaged liquor recognition method applicable to a multi-electronic nose platform. Background technique

[0002] The production and consumption of liquor plays an important role in the Chinese economy. With the development of the liquor market, the identification of liquor quality has attracted social attention. The commonly used liquor identification methods are divided into sensory evaluation, component analysis and electronic nose detection. Sensory evaluation is easily disturbed by external factors, while component analysis needs to rely on large precision instruments, which cannot be completed in real time. Compared with other analysis methods, the fast and low-cost characteristics of electronic nose make it widely used in the field of liquor identification.

[0003] Electronic nose technology, also known as artificial olfactory technology, is a method of odor detection, analysis and identification by imitating the str...

Images