不同光谱采集条件下麦秸低位热值NIRS快速检测

Rapid Detection of Net Calorific Value for Wheat Straw under Different NIRS Collection Conditions

采集我国不同地区、不同品种的71个麦秸样品,选取其中54个为校正集,17个为独立的验证集。采用Spectrum 400型和Antaris Target型两种近红外光谱仪,分别对麦秸低位热值进行实验室静态和模拟在线快速预测分析。利用Spectrum 400型近红外光谱仪对麦秸低位热值进行实验室静态分析,建立的预测模型校正集决定系数为0.90,验证集相对分析误差为2.75。利用Antaris Target型近红外光谱仪对麦秸低位热值进行实验室模拟在线分析,建立预测模型,校正集决定系数为0.93,验证集相对分析误差为3.26。对这两种光谱采集条件下的近红外光谱快速分析模型进行了比较分析。

A total of 71 wheat straw samples of different varieties were collected from different locations of China.54 samples were selected as calibration set,and the rest 17 samples were taken as validation set.Static and simulated online quantitative analysis models of net calorific value of wheat straw were established in laboratory by two different near-infrared spectrometers—Spectrum 400 and Antaris Target,respectively.Spectrum 400 was used for rapid detection in a static mode.The determination coefficient and the relative analysis error of calibration set were 0.90 and 2.75.Antaris Target was used to simulate on-line detection.The determination coefficient and the relative analysis error of calibration set were 0.93 and 3.26.Moreover,the static and simulated online models were compared and discussed.