Scholarlink

<p>During the drying process of Korla fragrant pear slices, achieving rapid, non-destructive moisture detection is crucial for precise process control and ensuring product quality. By using the 3671D vector network analyzer (VNA) in conjunction with the 9809 coaxial probe, the dielectric constant (ε’) and dielectric loss factor (ε”) of Korla pears at different moisture contents were systematically measured across 201 frequency points from 0.1 to 50 GHz, revealing their frequency-dependent patterns. The study found that within the 3.334 ~ 8.334 GHz band, ε’ exhibits a strong linear correlation with moisture content, whereas in the 14.571 ~ 36.527 GHz band, ε” shows a significantly stronger correlation with moisture content. Based on the selected characteristic frequencies, three moisture content prediction models were established: Principal Component Regression (PCR), Partial Least Squares Regression (PLSR), and Support Vector Regression (SVR). Results indicate that the model employing PLSR with ε” data from the 14.571 ~ 36.527 GHz band demonstrated optimal predictive performance, achieving a coefficient of determination (R²) of 0.899 and a root mean square error (RMSE) of 0.050. This study provides novel insights and methodologies for real-time moisture content prediction during fruit and vegetable drying processes and the development of related detection equipment.</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

Study on the moisture content of Korla fragrant pear based on dielectric properties

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Long Ouyang,
  • Guangxin Gai,
  • Kun Li,
  • Chen Ding,
  • Haipeng Lan,
  • Hong Zhang,
  • Jing An

摘要

During the drying process of Korla fragrant pear slices, achieving rapid, non-destructive moisture detection is crucial for precise process control and ensuring product quality. By using the 3671D vector network analyzer (VNA) in conjunction with the 9809 coaxial probe, the dielectric constant (ε’) and dielectric loss factor (ε”) of Korla pears at different moisture contents were systematically measured across 201 frequency points from 0.1 to 50 GHz, revealing their frequency-dependent patterns. The study found that within the 3.334 ~ 8.334 GHz band, ε’ exhibits a strong linear correlation with moisture content, whereas in the 14.571 ~ 36.527 GHz band, ε” shows a significantly stronger correlation with moisture content. Based on the selected characteristic frequencies, three moisture content prediction models were established: Principal Component Regression (PCR), Partial Least Squares Regression (PLSR), and Support Vector Regression (SVR). Results indicate that the model employing PLSR with ε” data from the 14.571 ~ 36.527 GHz band demonstrated optimal predictive performance, achieving a coefficient of determination (R²) of 0.899 and a root mean square error (RMSE) of 0.050. This study provides novel insights and methodologies for real-time moisture content prediction during fruit and vegetable drying processes and the development of related detection equipment.