• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Proceedings of joint international agricultural conference (JIAC 2009) >NON-DESTRUCTIVE MEASUREMENT OF SOLUBLE SOLID CONTENT IN PERSIMMON USING VISIBLE_NIR SPECTROSCOPY BASED ON PCA AND BP
【24h】

NON-DESTRUCTIVE MEASUREMENT OF SOLUBLE SOLID CONTENT IN PERSIMMON USING VISIBLE_NIR SPECTROSCOPY BASED ON PCA AND BP

机译：基于PCA和BP的VISIBLE_NIR光谱法对柿子中可溶性固形物的非破坏性测量

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In order to achieve fast and non-destructive measurement of soluble solid content (SSC) in persimmon( diospyros kaki thunb), a new method based on Visible_near infrared reflectance (NIR) spectroscopy was put forward. A Field Spec 3 spectroradiometer was used for collecting 22 sample spectra data of the three kinds of persimmon separately. Then principal component analysis (PCA) was used to process the spectral data after pretreatment using the average Smoothing method, and 6 principal components (PCs) were selected based on accumulative reliabilities. These selected PCs would be taken as the inputs of the three-layer back-propagation artificial neural network (BP-ANN). A total of 66 persimmon samples were divided into calibration sets including 51 samples（17 samples of each variety） and validation sets including 15 samples（5 samples of each variety） randomly. The three-layer BP-ANN model was established with 6 nodes that is 6 principal components (PCs) in input layer, 1 node that is soluble solid content (SSC) of persimmon in output layer and 11 nodes in hidden layer. Then this model was used to predict soluble solid content of persimmon for the sample in the validation set. The result showed that a standard error of calibration (SEC) of the calibration model was 0.232, and its prediction relative error below 3% was achieved, and the decision coefficient (R2) between the predicted value and the measurement value was 0.99, and the forecast standard deviation (SEP) was 0.257. It could be concluded that PCA combined with BP-ANN is an available method for soluble solid content measurement of persimmon based on NIR spectroscopy.
机译：为了实现柿果可溶性固形物含量的快速，无损检测，提出了一种基于可见-近红外反射光谱的新方法。使用Field Spec 3分光光度计分别收集三种柿子的22个样品光谱数据。然后，采用平均平滑法对经过预处理的光谱数据进行主成分分析（PCA）处理，并根据累积可靠性选择了6个主成分（PC）。这些选定的PC将作为三层反向传播人工神经网络（BP-ANN）的输入。柿子样品共66个，随机分为51个样品（每个品种17个样品）和15个样品（每个品种5个样品）的验证集。建立了三层BP-ANN模型，在输入层中有6个节点（即6个主要成分（PC）），在输出层中有1个节点是柿子的可溶性固形物（SSC），在隐藏层中有11个节点。然后使用该模型预测验证集中样品的柿子可溶性固形物含量。结果表明，校正模型的校正标准误差（SEC）为0.232，预测相对误差在3％以下，预测值与测量值的决策系数（R2）为0.99，预测标准差（SEP）为0.257。可以得出结论，PCA结合BP-ANN是一种基于NIR光谱测量柿子可溶性固形物含量的有效方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号