Research On Judging Method Of Rice Milling Degree With Near Infrared Diffuse Reflectance Spectroscopy

The milling degree of rice is an important parameter judging rice quality. The rice with higher milling degree shows not only better appearance, but also more appetizing and more welcome by consumers. However, the more milling time is, the more nutrimental component will be lost in the process. There is illogicality between milling degree and nutrimental quality. Therefore, prompt inspectation of the milling degree of rice is very important. At present, , the evaluation of milling degree of rice is conducted by manual method in the practical producing process in almost all country. Since this manual method of judging milling degree is subjective but objective, it is also difficult to realize the automatically control of rice milling. Therefore, an objective standard and inspecting method of assessing milling degree has been urgently required to guarantee milling quality and realize rice automatically milling.Previouse studies show that there is abvious difference in the nutrition contents between rice surface and its inter. The fat content of rice will decrease with the increase of milling degree. Therefore, the milling degree of rice can be assessed by the means of measuring fat content of rice.Due to advantages on analysis of farm products, the near infrared reflectance spectroscopy (NIR) is used in quantitative analysis and inspection of chemical composition content of farm products. The analysis technology can also be applied to the determination of fat content of rice. Moreover, the milling degree of rice can be judged according to the fat content and online examination of milling degree can be realized.The Vector22/N model FFT-NIR analytical instrument was utilized to measure the spectrum of rice fat content in this research. The NIDRS and PLS methods of chemometrics were adopted to establish mathematical model that can be used for the prediction of the fat content of rice surface. The relationships between the fat content and milling degree of rice were studied by combining the soxhlet fat extraction with GB of coloration and the believable region of fat content within per milling degree were confirmed. As a result, the milling degree was determined based on the region of the unknown samples' fat content.The Wu Yujing 3^rd was used as experimental materials. To begin with, simulating the process of rice milling, the experimental materials were processed to different milling degree by grinding lab-machine and milling lab-machine. The fat contents of calibration and validation samples were 0.160%～1.763% and 0.216%～0.927% respectively. One hundred and twenty calibration samples were analysed by using the auto-optimization function of the OPUS5.0 spectroscopic software. It was found that the optimal pretreatment method was multiplicative scattering correction (MSC), the optimal spectrum region from 11998.9cm^-1 to 4246.5cm^-1 and the optimal principal component 8. The cross validation was carried out and the calibration model with the above-mentioned parameters was established. The coefficient of determination (R²) for cross validation amounted to 0.9981 and the root mean square of error for the cross validation (RMSEC) amounted to 0.0191. The test set validation was carried out with 18 validation samples, whose true contents had been known, with the R² of cross validation 0.9889 and the root mean square of error for prediction (RMSEP) 0.0241. The correlation value (r) was 0.9960 and the relativity is significant. the true and the prediction were almost equal, meaning that the model could be practicably used to forecast fat contents of the unknown samples.If the milling degree was evaluated on the basis of the believable region of fat content within per milling degree, the correct rate was 83.33%, which indicated that the objective determination model of the milling degree could be established using the NIR analysis technology.Because the milling degree of rice was also distinguished by whitening rate of brown rice, if the NIR calibration model using whitening rate of brown rice was established and regarded as determination index for the milling degree of rice, that could avoid bringing in by the above-mentioned method, and the analysis process should be much more direct and detailed, the forecast effect absolutely reached the lab-require.Because of the advantage of online analysis and long-distance control, the NIR analysis technology had big potential in realizing online real-time determination of rice milling degree, and had quite important significance for guidance of online manufacturing, storage and quality study.