Studies Of Nondestructive Testing Technology To Beef Tallow Production Quality Control Based On Near Infrared Spectroscopy And Chemometrics Applications

Beef tallow is an important animal fat in the food industry.For example,in the baking and hot pot catering industry,the consumption is very large.The processing quality of the beef tallow is closely related to the storage and transportation methods,varieties,parts and processes of the raw materials.The processed cooking characteristics of its finished products are also related to the combination of different varieties of beef tallow.However,the quality control technology and measures in the existing beef tallow production industry are still too singular,and the extensive quality control of the beef tallow products is often followed by nonquantitative observation methods such as empirical process parameters and eye-catching such as water vapor bubbling,which is not conducive to the specification of beef tallow products of standardization development.In this study,near-infrared spectroscopy,gas chromatography,low-field NMR and other detection techniques were combined with chemometric methods to investigate the identification of raw material storage quality,identification of raw fat content,online non-destructive testing of processing and refining,and fatty acid composition differences.Quality control technology for identification and identification,Provide the modern quality control non-destructive testing ideas and research theoretical basis for the beef tallow production industry,and improve the ability and level of quality control of beef tallow production.The main conclusions are as follows:1.Taking the tripe oil of cattle(TOC)as the research object,the partial least-squares(PLS)analysis model was established by using the near-infrared spectroscopy data of TOC and acid value.By comparing the five parameters of RMSEC,Rc,RMSECV,RMSEP,Rp.The optimal scanning parameters for modeling the beef tallow using the portable NIR-500 NIR spectrometer from Chengdu Yiyang Modern Technology Co.,Ltd.were determined as follows: optical path 10 mm,scanning times 10 times,scanning temperature 30 °C.2.Taking the TOC raw materials under different storage conditions(Normal temperature(25°C);4?,-18?;-18°C repeated freezing)as the research object.Low-field nuclear magneticres-onance(LF-NMR)technique was used to study the moisture content and distribution under different storage conditions,and the near-infrared spectroscopy data was used for modeling.From the relaxation time spectrum,it can be concluded that the content of bound water,immobilized water and free water of the TOC under different storage conditions is different.From the MRI images of the samples under different storage conditions,it is found that the moisture distribution of the sample at normal temperature is not Uniform,with the gradual decrease of temperature,the binding ability of the sample system to moisture is gradually weakened,and the water distribution is gradually uniform.Using near-infrared spectroscopy data,SVM analysis and modeling can be used to identify the raw materials of yellow burdock under different storage conditions,and the predicted accuracy is 91.67%.3.Taking tripe oil of cattle(TOC),bovine oil of cattle(BOC),tripe oil of yak(TOY),bovine oil of yak(BOY)as research object,the variable adjustment scanning parameters of the portable near-infrared instrument were determined to be optical path 10 mm,scanning times 10 times,scanning temperature 30 °C.Then the above-mentioned beef tallow was labeled as 1,2,3,and 4 respectively,combined with the collected near-infrared spectral data,and modeled by PLS algorithm.The results showed that the correct rate of prediction for different types and different parts of beef tallow raw materials was found.83.3% can identify and classify the beef tallow raw materials of different types and different parts to a certain extent.4.Taking different finished tallows as the research object,the fatty acid detection of the finished tallow A,tallow B and tallow C was first carried out by gas chromatography,and then different fatty acids were regarded as multivariate factors to establish a PCA analysis model.It was found that PCA analysis can well classify different finished tallow;The near-infrared spectroscopy data of different finished tallows were collected by near-infrared spectroscopy.The classification model of different finished tallows was established by combining PLS analysis method.The prediction accuracy of this model is not 91.67%,which can predict the types of different finished tallows.5.The semi-finished tallow at different time points in the processing process was taken as the research object.Firstly,the moisture of the semi-finished tallow at different time points(0,5,10,15,20,25,30 min)during the tanning process was detected by a rapid moisture analyzer.The content of bovine oil content decreased with time,and gradually decreased and finally gradually stabilized;The near-infrared spectroscopy data of semi-finished tallow at different time points were collected,and the analytical model was established by SVM analysis method.The prediction accuracy of the model was found to be 78.6%,which had a good predictive effect on the endpoint judgment during the processing.