Research On Detection Methods Of Meat Freshness By Texture Analyzer

Food safety issues of meat related to the safety of human life directly, it’simportant to control food safety. However, conventional testing methods are unable tomeet the requirements of fast-paced life, the establishment of a rapid andnon-destructive detection method is particularly important. This paper aims todetermine the freshness of meat by Texture Analyzer, mainly as follows:1. Selected the fresh pork as a test，using the Kjeldahl nitrogen method for thedetermination of volatile basic nitrogen, The texture was determined by Brookfield’sCT3Texture Analyzer.Researched on the different effect of the texture indicators under differentexperimental conditions, just as the probe size, pre-stress, restore time, the test rateand compression deformation, and then determine the best conditions of TexturePro-file Analysis for meat, which proposes a probe diameter of12.7mm, the preloadis0.07N, the recoverability time is6s, the speed is4mm/s and compresseddeformation is5mm.2. Test pork (beef) using the recommended conditions, derived eight Textureparameters from the TPA curve: Hardness, compression work, restore function,recoverable deformation, resilience, elasticity, cohesion and corrected cohesion.Analyze the relevance of sensory evaluation and the determination ofinstrument，found that the relevance of the instrument determination and sensoryevaluation for hardness is0.965,and obtain the linear regression equation which Rsquare is0.931, but the relevance of elastic is poor. Determination the pork ninedays，put TVB-N as the freshness criteria, then analyze it with the texture indicators.Obtain the regression equation for the analysis of variance，proved that the textureindex and TVB-N has a larger correlation.3. Taking pork as the testing sample, using partial least squares elasticity toanalyze the regression equation between texture index and the freshness, Using it to predict the value of TVB-N, relative error is1.3%. In addition, we put in use the BPneural network between pork freshness and texture index and establish a nonlinearforecasting model, the detection accuracy rate is88.89%. The results of two model aresatisfactory, indicating that these can be applied to predict pork freshness.