| 神经网络在大米蛋白质含量预测模型中的应用 |
| |
| 引用本文： | 刘建学,吴守一,方如明.神经网络在大米蛋白质含量预测模型中的应用[J].江苏大学学报(自然科学版),2004,25(3):196-198. |
| |
| 作者姓名： | 刘建学 吴守一 方如明 |
| |
| 作者单位： | 1. 河南科技大学食品与生物工程学院,河南,洛阳,471003
2. 江苏大学生物与环境工程学院,江苏,镇江,212013 |
| |
| 基金项目： | 河南省自然科学基金资助项目(1347001) |
| |
| 摘 要： | 大米蛋白质含量是影响大米营养及食味特性的一个重要指标．借助主成分分析，确立了用于近红外光谱分析的BP神经网络的输入输出模式对；并用BP神经网络方法建立了不同类型、不同粒度的大米样品蛋白质含量预测模型；考察了模型的预测能力，其预测值与用标准方法取得的化学分析值间具有良好线性关系，相关系数达0．90以上，用BP神经网络可降低样品粒度的不同对预测结果造成的差异．
|
| 关 键 词： | 大米 蛋白质含量 近红外光谱 BP神经网络 预测模型 |
| 文章编号： | 1671-7775(2004)03-0196-03 |
| 修稿时间： | 2003年11月16 |
Determination of protein content of rice by near infrared spectroscopy based on neural networks |
| |
| LIU Jian-xue,WU Shou-yi,FANG Ru-ming.Determination of protein content of rice by near infrared spectroscopy based on neural networks[J].Journal of Jiangsu University:Natural Science Edition,2004,25(3):196-198. |
| |
| Authors: | LIU Jian-xue WU Shou-yi FANG Ru-ming |
| |
| Institution: | LIU Jian-xue~1,WU Shou-yi~2,FANG Ru-ming~2 |
| |
| Abstract: | Protein content (PC) is one of the important parameters in affecting the nutrition and taste characteristics. The chemical measurement of rice PC is expensive and time consuming. The input pattern of BP neural networks based on near infrared spectroscopy and principal component regression is presented, and the BP neural networks (BPNN) model of PC for the rice samples of different particles and varieties is built by optimizing the studying vector. The correlation coefficient between the evaluated value of rice PC by BPNN algorithm and chemical value is 0.94. It shows that the BP neural networks can reduce the predicting difference produced by the rice samples of different particles. |
| |
| Keywords: | rice protein content near infrared spectroscopy BP neural networks determination model |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
