Real-time Monitoring The Oxygen Process Of Summer And Fall Tea Extract Using Visible/near Infrared Spectroscopy

Oxygenation of summer and fall tea extract can reduce its bitterness and turn it into instant black tea, thus will improve the utilization rate of summer and fall tea. However, whether we can assess the degree of the oxygenation process is the key to the quality of instant black tea. At present, most of the factories evaluate the degree of the oxygenation process of summer and fall tea extract by observing its color change, and then verify it by measuring tea polyphenol content via the traditional chemical analytical methods. Nonetheless, there are many shortcomings of these methods, such as strong subjectivity, poor repeatability and low real-time performance,which affect the stability of product quality. Hence, a research of real-time monitoring the oxygen process of summer and fall tea extract using visible/near infrared spectroscopy was carried out, and a set of real-time monitoring system was made. The main research content is as follows:1. The design of the real-time monitoring system of summer and fall tea extract. The real-time monitoring system consists of hardware system and software system. The hardware system consists of spectrometer, light source, optical fiber, sample pool, embedded terminal, PC,gas-liquid mixing tube, storage tank, air pump, liquid pump and micro peristaltic pump. The wavelength range of the spectrometer is 300-1000 nm; the type of light source is halogen lamp;the material of optical fiber is quartz; the sample pool used for real-time access to liquid samples consists of flowing colorimetric dish, collimating lens and external stent; the embedded terminal is based on Samsung S3C2440; the gas-liquid mixing tube is made by a spiral circle pipe; the storage tank is used to store the summer and fall tea extract; the air pump is used to carry oxygen;the liquid pump is used to extract summer and fall tea extract to be mixed; the micro peristaltic pump is used to take summer and fall tea extract to sample pool. Software system mainly refers to the embedded terminal and PC software. The embedded terminal software consists of Boot Loader, Linux kernel, Qt Embedded and the application consisting of communication module and user interface. The PC software consists of communication module, spectrum processing module and user interface. This monitoring system can acquire real-time visible/near infrared spectra of summer and fall tea extract in the process of oxygenation to calculate chromaticity values and tea polyphenols content.2. Real-time monitoring the chromaticity values of summer and fall tea extract in the process of oxygenation. Firstly, we monitored the chromaticity values of summer and fall tea extract during the process of oxygenation, and then determined the degree of oxygenation based on the change in its chromaticity values. To begin with, we get integer wavelengths spectralintensity from the original transmission spectrum by data fitting, subsequently calculate stimulus values by equal wavelength interval method to get L*a*b* values. The calibration of monitoring system was endured with standard colorimetric fluid and color difference meter, afterward, the accuracy of color measurement model was up to 99%. In consideration of the calibrated color measurement model, we test its accuracy with summer and fall tea extract of different degree of oxygenation for samples compared with the results of standard color difference meter, and monitor the change in chromaticity value of summer and fall tea extract in the process of oxygenation. In order to check the feasibility of real-time monitoring the oxygenation process of summer and fall tea extract based on chromaticity value, we also tested its accuracy with summer and fall tea extract of different degree of oxygenation for samples compared with the results of artificial expert system, as a result, the accuracy rate of monitoring system was up to96.25%. In conclusion, it is feasible to monitor the chromaticity value of summer and fall tea extract in the process of oxygenation using visible/near infrared spectrum and judge the degree of oxygenation process of summer and fall tea extract based on chromaticity value.3. Real-time monitoring the polyphenol content of summer and fall tea extract in the process of oxygenation. Initially, we monitor the polyphenol content of summer and fall tea extract in the process of oxygenation, and then determine its degree of oxygenation according to the change in their respective polyphenol content. We establish a quantitative prediction model of polyphenol content and visible/near infrared spectrum of summer and fall tea extract in the process of oxygenation. Then, we acquire the visible/near infrared transmission spectrum, and simultaneously measure the polyphenol content using traditional chemical analysis method. Next,we preprocess the spectrum using SNV method, and then build model using synergy interval partial least square(Si-PLS), genetic algorithm-partial least squares(GA-PLS), competitive adaptive reweighted sampling(CARS) and Ant Colony Optimization(ACO) comparatively. As a result, the number of variables, correlation coefficient of prediction(Rp) and root mean squared error of prediction(RMSEP) using full spectrum PLS method were 902, 0.7659 and 0.0707respectively; the number of variables, Rp and RMSEP using Si-PLS method were 212, 0.8766 and 0.0535 respectively; the number of variables, Rp and RMSEP using GA-PLS method were145, 0.8993 and 0.05 respectively; the number of variables, Rp and RMSEP using CARS method were 65, 0.8897 and 0.0517 respectively; the number of variables, Rp and RMSEP using ACO method were 20, 0.8853 and 0.0508. The results showed that all the four different methods were better than the full spectrum PLS method. Similarly, the ACO was the best method considering the circumstances, so the ACO model was selected as the best quantitative prediction model of polyphenol content. In order to check the feasibility of real-time monitoring the oxygenationprocess of summer and fall tea extract based on polyphenol content, we test its accuracy with summer and fall tea extract of different degree of oxygenation for samples compared with the results of artificial expert system, as a result, the accuracy rate of monitoring system was up to98.5%. Conclusively, it is highly practicable to monitor the polyphenol content of summer and fall tea extract in the process of oxygenation using visible/near infrared spectrum and judge the degree of oxygenation process of summer and fall tea extract based on polyphenol content.