【作者】 张彬；

【导师】 陈全胜；

【作者基本信息】 江苏大学 ， 食品科学与工程， 2017， 硕士

【摘要】 发酵是红茶制作中最重要的一道工艺,其本质是以茶多酚为核心的生理生化反应,并伴随着叶片表面色泽的显著变化,形成独具风格的色香味等特征品质,也是决定红茶质量的关键。目前大部分红茶加工企业仍然依靠有经验的制茶师傅通过“一看二闻”来判断发酵是否适度。但此法易受环境因素干扰、主观性大、效率低,无法做到红茶加工过程中的自动化和连续性,难以保证红茶品质的稳定性。基于此,本论文展开了基于光学传感器技术的红茶通氧发酵过程在线监测研究。主要研究内容与结论如下:1.红茶通氧发酵过程在线监测系统的设计。在现有的红茶通氧发酵设备的基础上,设计开发了两套在线监测设备,分别用于监测通氧发酵过程中叶片表面色泽和内部茶多酚含量的改变,均包括硬件和软件系统。硬件主要包含SD1220和集成近红外光谱仪、光源、积分球、光纤、样品杯、蓝牙模块以及上位机。其中,SD1220和集成近红外光谱仪的有效波长范围分别为300~1000 nm、900~1700nm;光源是卤钨灯;积分球为反射式;直通光纤的材质是石英;样品杯为定制的圆型、无盖、底部1mm厚的石英;蓝牙模块的型号包括HC-06和CC2564MODN;上位机为Android智能机。软件主要包括蓝牙通信、光谱数据处理和结果显示以及客户端界面设计模块。其中,蓝牙通信模块主要包含查找光谱仪设备,并连接通讯;光谱数据处理和结果显示模块主要包括前处理、计算L*a*b值或茶多酚含量以及显示模型判别结果;客户端界面设计模块主要包括一些操作按钮与图形显示界面,用于连接蓝牙、采集光谱、计算结果及保存显示等功能。这两套在线监测系统分别用于红茶通氧发酵过程中的可见与近红外光谱的采集,并计算叶片的L*a*b值和茶多酚含量。2.红茶通氧发酵过程中叶片色度值的在线监测研究。研究应用可见漫反射光谱技术在线监测红茶通氧发酵过程中叶片外部的色泽,并依据L*a*b值的改变实时评判叶片的发酵程度。首先,对发酵茶样的原始反射光谱进行处理,拟合运算以获取380~780nm下的整波长数值。结合GB/T3979-2008的测色原理,采用等波长间隔法计算XYZ三刺激值,并换算成L*a*b值。其次,对该仪器的测色准确性进行验证。以杭州茶叶研究所提供的三种不同等级红茶标品为样本,利用该系统和标准色差计同时测量,所得L*a*b值回归方程的相关系数均高于0.9,表明此仪器的精度与稳定性较高。然后,构建发酵适度的色泽评判模型。收集经感官评审小组确定的发酵适度时间节点下的60个样本,计算L*a*b值并建立判别模型。最后,验证此模型用于在线监测红茶通氧发酵程度的可行性。以60个不同发酵程度的红茶为验证样品,利用本监测系统进行检测,同时结合人工评审结果,正确判别率为85%。研究结果证明,基于叶片颜色值在线监测评判红茶通氧发酵程度是可以的。3.红茶通氧发酵过程中茶多酚含量的在线监测研究。研究利用近红外光谱技术在线监测红茶通氧发酵过程中叶片的茶多酚含量,并依据其值变动实时评判叶片的发酵程度。首先,处理原始光谱并构建红茶通氧发酵过程中茶多酚总量的定量回测模型。以不同时间节点的发酵叶片为样本,采集其反射光谱,再根据GB/T8313-2008第二法检测茶多酚总量。随后利用SNV进行前处理后,分别采用全波段、联合区间(SI)、遗传算法(GA)、连续投影算法(SPA)和蚁群算法(ACO)结合PLS进行变量筛选与建模分析。结果表明,相比于全光谱模型,其余四类变量筛选方法均可有效提高模型预测性能,并且ACO-PLS模型预测效果与稳定性最好,该模型的有效变量为19个,Rp=0.8317,RMSEP=1.15。因而,选取ACO-PLS作为茶多酚含量的最佳回测模型。其次,建立发酵适度的茶多酚含量评判标准。采集经感官评审小组确认的发酵适度时间节点下的60个样本,使用国标测定茶多酚总量并制订判别标准。最后,验证该模型用于在线监测红茶通氧发酵过程的可行性。以60个不同发酵程度的红茶叶片为验证样本,利用本监测系统进行测定,并结合人工评审结果,正确判别率为81.67%。研究结果证明,基于叶片茶多酚含量变化在线监测评判红茶通氧发酵程度是可行的。更多还原

【Abstract】 Fermentation is the most important processing step of the production of black tea,and its essence is a series of physiological and biochemical reactions based on tea polyphenols,following with the significant changes on the surface of leaf surface and the generation of unique characteristic qualities which includes color,fragrance and flavor.Significantly,it is also the critical factor to determine the quality of black tea.At present,most of the enterprises engaged in black tea processing still rely on experienced tea makers who adopt the method of ’looking and then smelling’ to determine whether the fermentation degree is appropriate.However,this method is susceptible to environmental factors,subjective,has low efficiency and lacks automation and continuity during black tea production process.This therefore makes it difficult to ensure the stability of black tea quality.Sequel,this study developed an on-line optical sensor device for monitoring aerobic fermentation process of black tea.The main research contents and conclusions are as follows:1.Design of on-line monitoring system for aerobic fermentation process of black tea.On the basis of the existing black tea aerobic fermentation equipment,two sets of on-line monitoring equipment which consisted of hardware and software systems,were designed and used to monitor the changes of leaf surface color and internal tea polyphenol content during the aerobic fermentation process separately.Hardware mainly included SD1220 and integrated near infrared spectrometer,light source,integrating sphere,optical fiber,sample cup,bluetooth module and the host computer.Among them,the effective wavelength of SD1220 and integrated near infrared spectrometer were in the range of 300-1000 nm and 900-1700 nm respectively;light source was halogen lamp;integrating sphere was reflective;the material of straight optical fiber was quartz;the sample cup was round with no cover,and had 1mm thick quartz bottom;the bluetooth module models included HC-06 and CC2564MODN;and an android smart phone was employed as the host computer.The software mainly included bluetooth communication,spectral data processing and result display and client interface design module.Among them,the bluetooth communication module mainly included the signal locator of the spectrometer equipment and connector of the communication;the spectral data processing and the result display module mainly included the pre-processing,the calculation of the L*a*b values or the tea polyphenol content and the display model judgment result;the client interface design module mainly included some operation button and graphical display interface for connecting bluetooth,acquisition of the spectrum,computed outcomes and save display,and other functions.The two sets of on-line monitoring system were used for the collection of visible and near infrared spectra during the process of black tea aerobic fermentation and the calculation of leaf L*a*b values and tea polyphenol content.2.Study on the on-line monitoring of the chromaticity of leaf during the aerobic fermentation of black tea.The visible diffuse reflectance spectroscopy was used to monitor the color of the leaves outside during the aerobic fermentation of black tea,and the fermentation degree of the leaves were judged in real time according to the change in the L*a*b values.Firstly,the original reflection spectrum of the fermented tea samples were processed and fitted to obtain the whole wavelength values between 380 ~ 780 nm.According to the colorimetric standards of China(GB / T 3979-2008),the XYZ tristimulus values were calculated by the equal wavelength interval method,and converted into L*a*b values.Secondly,the color measurement accuracy of the instrument was verified.Three different levels of black tea,obtained from Hangzhou Tea Research Institute,were used as the standard samples;and the developed system and standard colorimeter were measure them at the same time.The correlation coefficients of regression equation of L*a*b values were higher than 0.9,which showed high accuracy and stability of the developed instrument.Then,a moderate color evaluation model was constructed.A total of 60 samples from the appropriate time node of the fermentation process was determined by the sensory evaluation group were judged as good,and then calculated L*a*b values were used established the discriminant model.Finally,the feasibility of the application of this model for on-line monitoring of the degree of aerobic fermentation of black tea was verified.The correct judgment rate was 85% by using 60 kinds of black tea with different fermentation degree as the verification sample based on the monitoring system and the sensory panel review results.The results showed that it is possible to judge the degree of oxygenation of black tea by online monitoring based on leaf color values.3.Study on the on-line monitoring of tea polyphenol content during the aerobic fermentation of black tea.The tea polyphenol content in the leaves of black tea during the aerobic fermentation was monitored by near infrared spectroscopy,and the fermentation degree of the leaves was evaluated in real time according to the change of the values.Firstly,the quantitative prediction model of the total amount of tea polyphenols during the process of oxygenation was established by processing the original spectrum.The reflectance spectra of the samples were collected at different time points,and the total amount of tea polyphenols was determined according to Chinese standards,second alternative method(GB / T 3979-2008).The SNV algorithm was used to preprocess the raw spectrum,and then the variable selection methods,including the full band interval,synergy interval(SI),genetic algorithm(GA),successive projections algorithm(SPA)and ant colony algorithm(ACO),were combined with PLS respectively to select variables and build the model.The results showed that compared with the full spectrum model,the other four types of variable screening methods effectively improved the model performance,and the prediction model obtained by ACO-PLS algorithm had the best result and the highest stability.The effective variables of the model were 19,Rp = 0.8317,RMSEP = 1.15.Therefore,ACO-PLS algorithm was selected as the best prediction model of tea polyphenol content.Secondly,the model of appropriate fermentation of tea polyphenols content was established.A total of 60 samples were collected under the appropriate fermentation time node identified by the sensory evaluation team and the national standard was used to determine the total amount of tea polyphenols,then the criterion was established.Finally,the model was validated for the on-line monitoring of the feasibility of the black tea oxygenation process.The correct discrimination rate of 81.67% was obtained by testing 60 samples of black tea leaves at different fermentation levels using the monitoring system and the manual review result.The results showed that it is feasible to evaluate the degree of tea polyphenols based on the change of tea polyphenols content.更多还原