| 马强,李水芳,付红军,王琼,文瑞芝.油桐籽不同年份和含油率差异对其含油率NIR检测模型影响的研究[J].中国粮油学报,2019,34(6):92-96 |
| 油桐籽不同年份和含油率差异对其含油率NIR检测模型影响的研究 |
| Assessment of Influence of Year and the Oil Content Variability on Robustness of Near Infrared Models for Oil Content of Vernicia fordii Seeds |
| 投稿时间：2018-08-18 修订日期：2018-11-16 |
| DOI： |
| 中文关键词: 油桐籽 年份 含油率 近红外光谱 |
| 英文关键词:Vernicia fordii Seeds year oil content Near-infrared spectroscopy |
| 基金项目:湖南省教育厅重点项目（14A155） |
|
| 摘要点击次数: 645 |
| 全文下载次数: 27 |
| 中文摘要: |
| 收集了2014和2015年的油桐籽样本，用偏最小二乘法（PLS）分别建立了单一年份、混合年份及单一年份不同含油率范围的油桐籽含油率近红外光谱（NIR）检测模型，并验证。结果显示单一年份模型对本年份样本有较好预测，而对另一单一年份样本的预测精度明显下降，而混合年份模型对各年份样本都有较好预测；同年份不同含油率范围样本所建模型，含油率范围大，则模型预测精度下降，但稳定性更好。利用竞争性自适应重加权（CARS）算法筛选出30个变量，并结合PLS对混合年份样本建模，既简化了模型，又提高了模型预测性能，验证集相关系数为0.929，均方根误差为1.765，相对标准偏差为3.31%。因此，建立油桐籽含油率NIR检测模型时，应收集不同年份、不同含油率范围样本，并结合特征波长，以建立预测精度更好、稳定可靠且适应范围广的检测模型。 |
| 英文摘要: |
| s: The samples of Vernicia fordii Seeds from 2014 to 2015 were collected. Partial least squares (PLS) was used to develop NIR calibration models for oil content of Vernicia fordii seeds by the samples from particular year, hybrid years and the samples of the different oil content range from the same year, respectively, and then the models were used to predict the different validation samples. The results showed that the models developed by the samples from particular year could be effectively applied to the samples from same year but not another year, however the models developed by the samples from hybrid years could effectively applied to all the samples and the models developed by the samples of the larger oil content range showed worse prediction, but better robustness than that developed by the samples of the smaller range. The model could not only be effectively simplified but also be taken for improvement using 30 characteristic variables selected by Competitive adaptive reweighted sampling (CARS). The correlation coefficient (Rp), root mean square error (RMSEP) and relative standard deviation (RSDp) of validation set were 0.929, 1.765 and 3.31%, respectively for oil content. It was found that the robust and effective model of oil content of Vernicia fordii seeds could be developed by more samples of different year and larger oil content range combined with characteristic variables. |
| 关闭 |
