Scholarlink

<p>Currently, the wet-chemical analysis method is primarily used to detect the components of hemp fiber. However, this method is time-consuming and not environmentally friendly. This paper presents a study on the detection of the main components of hemp using near-infrared (NIR) spectroscopy and their determination through wet chemical analysis. The relationship between chemical analysis data and NIR spectral data was established using the partial least squares (PLS) and principal component regression (PCR) methods. Based on the corrected and predicted root mean square error (RMSE) and mean absolute error (MAE), it can be concluded that PCR is a more effective quantitative method than PLS. The constructed main component regression prediction models for cellulose, hemicellulose, and lignin had RMSE values of 2.24%, 0.83%, and 1.87%, respectively, while their MAE values were 5.89%, 8.21%, and 2.24%. These results indicate good stability. Optimizing the spectral wavelength range improved the modeling quality of the PCR prediction model for cellulose, hemicellulose, and lignin. The spectral wavelength range for cellulose was 1600–2400&#xa0;nm, with RMSE and MAE of 3.78% and 3.88%, respectively. The spectral wavelength range for hemicellulose was 1400–2400&#xa0;nm, with RMSE and MAE of 1.37% and 14.56%, respectively. The spectral wavelength range for lignin was between 1200 and 2400&#xa0;nm, with RMSE and MAE of 3.03% and 17.79%, respectively. These results demonstrate that the NIR model offers a quick and straightforward approach to detecting components in hemp fiber, which is beneficial for evaluating its quality.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Rapid Determination of Cellulose, Hemicellulose and Lignin Content in Hemp Fibers Using Near-Infrared Spectroscopy

  • 期刊论文

  • 发表日期：

    2025年2月10日

查看全文
  • Lijian Wang,
  • Chao Lu,
  • Jiangang Wang,
  • Chunhong Wang,
  • Cuiyu Li

摘要

Currently, the wet-chemical analysis method is primarily used to detect the components of hemp fiber. However, this method is time-consuming and not environmentally friendly. This paper presents a study on the detection of the main components of hemp using near-infrared (NIR) spectroscopy and their determination through wet chemical analysis. The relationship between chemical analysis data and NIR spectral data was established using the partial least squares (PLS) and principal component regression (PCR) methods. Based on the corrected and predicted root mean square error (RMSE) and mean absolute error (MAE), it can be concluded that PCR is a more effective quantitative method than PLS. The constructed main component regression prediction models for cellulose, hemicellulose, and lignin had RMSE values of 2.24%, 0.83%, and 1.87%, respectively, while their MAE values were 5.89%, 8.21%, and 2.24%. These results indicate good stability. Optimizing the spectral wavelength range improved the modeling quality of the PCR prediction model for cellulose, hemicellulose, and lignin. The spectral wavelength range for cellulose was 1600–2400 nm, with RMSE and MAE of 3.78% and 3.88%, respectively. The spectral wavelength range for hemicellulose was 1400–2400 nm, with RMSE and MAE of 1.37% and 14.56%, respectively. The spectral wavelength range for lignin was between 1200 and 2400 nm, with RMSE and MAE of 3.03% and 17.79%, respectively. These results demonstrate that the NIR model offers a quick and straightforward approach to detecting components in hemp fiber, which is beneficial for evaluating its quality.