| The crude fiber,moisture and ash content of tea were closely related to tea quality.At present,the physical and chemical detection methods of crude fiber,moisture and ash in tea have the advantages of mature method and high accuracy,but the process of the treatment was complicated,the consuming of time was large,the cost was high,the damage of sample was large,the reagent consumption was large,and the special equipment was needed,which limited the detection efficiency.Compared with the traditional physical and chemical detection method,the near infrared spectroscopy helps to improve the detection efficiency,reduce the personnel costs,have no using of chemical reagents,have no damage to the sample,which is a way of green environmental protection test.In the study,near infrared spectrum analysis technique was adopted,and the quantitative analysis of crude fiber,moisture and ash content in tea was carried out by combining with the method of chemometrics.The main research contents and conclusions are as follows:(1)In this study,one hundred and thirty five tea samples were randomly selected and their crude fiber,moisture and ash contents were determined by traditional physical and chemical detection methods.The results show that the crude fiber content in all the samples is between 7%~18%,and the samples with the fiber content between 9% ~ 13% is account for 76.3% of all the samples.The water content is between 2% and 9%,and the samples with the water content between 5% and 8% account for 94.8% of the all the samples.The ash content is between 4% and 9%,and the samples with the ash content between 5 % and 6 % account for 80.7% of the samples.(2)The optimum spectral acquisition parameters of the samples were determined by experiments: thickness of 4 mm,scanning times of 32 times and particle size of 0.5 mm.Under the above parameters,one hundred and thirty five tea samples were scanned by near infrared spectroscopy,and the mathematical model was established by combining the obtained map data with chemometrics.In the model evaluation,The correlation coefficients(r)of crude fiber calibration set and validation set were 0.9543 and 0.9623 respectively,and the relative analysis error(RPD)was 3.49 and 3.56 respectively,The correlation coefficient(r)of that water correction set and the verification set were 0.9892 and 0.9868 respectively,and the relative analysis error(RPD)was 3.27 and 3.38 respectively,The correlation coefficients(r)of ash correction set and verification set were 0.9815 and 0.9838,and the relative analysis errors(RPD)were 3.49 and 3.17,respectively.The correlation coefficient(r)was above 0.90,and the linear correlation was good.The relative analysis error(RPD)was greater than 3,and the prediction accuracy of the model was high.(3)The established quantitative analysis model of near infrared spectroscopy was verified externally.The near infrared model was used to predict the 60 samples which have not participated the establishment of the model.Near infrared spectroscopy and traditional physical and chemical detection method were carried out by the test and get the content of the crude fiber,moisture and ash of the sample in t value were 1.26,1.37,1.65,less than the critical value of t(0.05,60)= 2.00,which shows that there is no significant difference between the two methods,and the consistency is better.(4)The results showed that near infrared spectroscopy could be used to quickly detect the contents of crude fiber,moisture and ash of tea,and laid a foundation for the wide application of near infrared spectroscopy in food and other fields. |
PDF Full Text Request