为了鉴别西湖龙井和浙江龙井茶叶,采用近红外光谱分析技术结合化学计量学方法建立了识别模型。先对原始光谱进行标准正态变换(Standard Normal Variant,SNV)预处理,然后分别采用最小二乘判别分析(Partial Least Square Regression-discr...为了鉴别西湖龙井和浙江龙井茶叶,采用近红外光谱分析技术结合化学计量学方法建立了识别模型。先对原始光谱进行标准正态变换(Standard Normal Variant,SNV)预处理,然后分别采用最小二乘判别分析(Partial Least Square Regression-discriminantAnalysis,PLS-DA)、最小二乘支持向量机(Least Square Support Vector Machine,LSSVM)和径向基人工神经网络(Radial Basis Function Neural Network,RBFNN)三种模型对西湖龙井和浙江龙井茶叶进行预测。最小二乘支持向量机参数通过网格搜索和完全交叉验证得到优化。经优化后,惩罚系数(γ)和核函数参数(δ~2)分别为229.1和124.9;RBFNN最佳隐藏层神经元个数为27个。通过比较可知,LSSVM的预测性能最好,其校正集均方根误差(RMSECV)和相关系数(R^2)分别为0和1,验证集均方根误差(RMSEP)和相关系数(R^2)也分别为0和1,分辨正确率为100%。展开更多
To understand the mineral elements in different varieties of tea （ Camellia sinensis）, 17 mineral elements in eight tea varieties including Yunnan Dayezhong, No. 43 Longjing and No. 6 Zaobaijian, were measured. The ...To understand the mineral elements in different varieties of tea （ Camellia sinensis）, 17 mineral elements in eight tea varieties including Yunnan Dayezhong, No. 43 Longjing and No. 6 Zaobaijian, were measured. The results showed that nine elements, such as P, K, Ba, Mn, Cu, were significantly different among varieties, others did not. Black tea varieties usually contain high contents of Cu and K. As a conclusion, mineral elements should be extensively considered in breeding tea varieties.展开更多
As one of the top ten Chinese teas, Longjing Green Tea is famous for its green tea leaves, sweet taste, pleasant aroma and beautiful shape. Besides of using as drinks, Longjing tea leaves can be applied as spice in Ch...As one of the top ten Chinese teas, Longjing Green Tea is famous for its green tea leaves, sweet taste, pleasant aroma and beautiful shape. Besides of using as drinks, Longjing tea leaves can be applied as spice in Chinese cuisine for its particular taste and function. Under the background of pursuing healthy cuisine, the development of organic Longjing tea and advocate of Chinese tea culture have great economic and culture significance.展开更多
文摘为了鉴别西湖龙井和浙江龙井茶叶,采用近红外光谱分析技术结合化学计量学方法建立了识别模型。先对原始光谱进行标准正态变换(Standard Normal Variant,SNV)预处理,然后分别采用最小二乘判别分析(Partial Least Square Regression-discriminantAnalysis,PLS-DA)、最小二乘支持向量机(Least Square Support Vector Machine,LSSVM)和径向基人工神经网络(Radial Basis Function Neural Network,RBFNN)三种模型对西湖龙井和浙江龙井茶叶进行预测。最小二乘支持向量机参数通过网格搜索和完全交叉验证得到优化。经优化后,惩罚系数(γ)和核函数参数(δ~2)分别为229.1和124.9;RBFNN最佳隐藏层神经元个数为27个。通过比较可知,LSSVM的预测性能最好,其校正集均方根误差(RMSECV)和相关系数(R^2)分别为0和1,验证集均方根误差(RMSEP)和相关系数(R^2)也分别为0和1,分辨正确率为100%。
基金Supported by Applied Basic Research Development Program of Si-chuanKey Scientific and Technological Project of Tea Breeding Pro-ject in Sichuan Province during Eleventh Five-year Plan~~
文摘To understand the mineral elements in different varieties of tea （ Camellia sinensis）, 17 mineral elements in eight tea varieties including Yunnan Dayezhong, No. 43 Longjing and No. 6 Zaobaijian, were measured. The results showed that nine elements, such as P, K, Ba, Mn, Cu, were significantly different among varieties, others did not. Black tea varieties usually contain high contents of Cu and K. As a conclusion, mineral elements should be extensively considered in breeding tea varieties.
文摘As one of the top ten Chinese teas, Longjing Green Tea is famous for its green tea leaves, sweet taste, pleasant aroma and beautiful shape. Besides of using as drinks, Longjing tea leaves can be applied as spice in Chinese cuisine for its particular taste and function. Under the background of pursuing healthy cuisine, the development of organic Longjing tea and advocate of Chinese tea culture have great economic and culture significance.
