利用近红外光谱技术快速测定木材水分和气干密度的研究

The study of rapid prediction of moisture content and air-dried density of wood by using near infrared spectroscopy technology

利用近红外光谱技术对木材水分和气干密度进行了快速测定。结果表明:两个水分模型的决定系数都接近1,RMSECV值小于0.2%,RPD值大于10,模型质量极好,对样品的预测偏差小于0.2%;气干密度模型的决定系数为0.976、RM-SECV值为0.0152g/cm3、RPD值为6.47,模型质量好,对样品的预测偏差范围为-0.019～0.02g/cm3。说明可以利用近红外光谱技术对我国造纸木材的水分和气干密度进行快速、准确的预测。

Rapid prediction of moisture content and air-dried density of wood by using near infrared spectroscopy was studied in this paper.Results show that the determination coefficients of 2 moisture contents models are close to 1, the RMSECV values are lower than 0.2%, the RPD values are greater than 10, the quality of the calibration models is exceedingly good, the prediction error is lower than 0.2%.The results of calibration model of air-dried density show that high determination coefficient is 0.976, the RMSECV value is 0.0152g/cm^3, and the RPD value is 6.47, the quality of air-dried density model is good, the prediction error ranges are-0.019～0.02g/cm^3.The results from this study show that moisture content and air-dried density can be rapidly, accurately predicted by NIR analytical technology.