Pork Freshness Detection System Based On Electronic Nose

In recent years, with the improvement of living standard, people’s food were required fromthe quantity stage to qualitative stage, food safety was paid more attention. Pork was one of thefoods that Chinese liked to eat; pork freshness was a main index reflecting the pork quality andsafety, so the freshness detection of pork became a very meaningful task. The traditional porkfreshness determination methods were sensory identification, physical, chemical, and microbialappraisal and so on, these methods had shortcomings such as long experiment cycle, tedioussteps, subjective factors. The purpose of this paper was to find fast, nondestructive and low inprice technology to detect pork freshness.Three main parts was carried out in this subject. The first part, using headspace solid phasemicro-extraction–gas chromatography–mass spectrometry, was determination of volatilecomponents in the pork stored at different temperature at5℃. The experiment showed that33kinds compounds were identified, including two kinds of nitrogen compounds, two kinds ofsulfur compounds, two kinds of ketone compounds,three kinds of alcohols, seven kinds ofaldehydes, four kinds of esters,13kinds of hydrocarbons. Pork sample stored at25℃showedthat39kinds compounds were identified, including three kinds of nitrogen compounds, fivekinds of sulfur compounds, two kinds of ketones compounds, three kinds of alcohols, sevenkinds of aldehydes, four kinds of esters,15kinds of hydrocarbons. The relative percentage ofNitrogen-containing, sulfur compounds and alcohols was increasing with storage time increasein the same storage temperature. The content of volatile components was higher at25℃than5℃in the same storage time; Total volatile basic nitrogen (TVB-N) content was increasing with theincrease of the temperature and time. In the detection of pork freshness using electronic noseexperiment, the output signals of three sensors in the sensor array was increasing with theextension of storage time, the three sensors were TGS825, TGS880, TGS826. Two models wereestablished to show the relationship of the output signal of the electronic nose and the value ofvolatile components using multiple linear regression analysis: one model was establishedaccording to the six-sensor output signal and the value of volatile components. The other wasestablished using the elected three sensor output signal and the value of volatile components.therecognition rate of the two models was respectively87.5%,81.25%. The recognition rate of thetwo models was little difference, so the model using the array of the elected three sensors wasfeasible to detect the freshness of the pork.