Scholarlink

<p>Moisture content is a critical indicator determining corn’s storage stability, processing performance, and trade value, which makes accurate and rapid measurement essential for optimizing the corn supply chain. This study proposes a non-destructive measurement method based on microwave technology. The experimental setup comprises a microwave pseudo waveguide and transceiver probe antennas. Scattering parameters were used to characterize the system and collected via a vector network analyzer. By analyzing the interaction between microwaves and corn samples, scattering parameters were correlated with moisture content data obtained from the traditional gravimetric method. A predictive model was developed using an autoencoder and multilayer perceptron regression algorithms, achieving excellent performance with a coefficient of determination of 0.9832, root mean square error of 0.0103, and mean absolute error of 0.0088. This method exhibits non-destructive, non-contact, and rapid measurement capabilities, suitable for real-time and online applications, thereby contributing to the advancement of intelligent management in agriculture.</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

Moisture Content Measurement System for Corn Grains based on Microwave Signal and Machine Learning

  • 期刊论文

  • 发表日期：

    2025年11月8日

查看全文
  • Huan Cai,
  • Yuqiu Yang,
  • Junyao Wu,
  • Miao Zhang,
  • Nianxing Hou,
  • Zixuan Guo,
  • Tao Zhou,
  • Wenqing Huang,
  • Qiaoling Sun,
  • Liangbin Deng,
  • Geng Jiang,
  • Bangyong Yin,
  • Xi Jiang,
  • Jungang Yin,
  • Linfeng Deng

摘要

Moisture content is a critical indicator determining corn’s storage stability, processing performance, and trade value, which makes accurate and rapid measurement essential for optimizing the corn supply chain. This study proposes a non-destructive measurement method based on microwave technology. The experimental setup comprises a microwave pseudo waveguide and transceiver probe antennas. Scattering parameters were used to characterize the system and collected via a vector network analyzer. By analyzing the interaction between microwaves and corn samples, scattering parameters were correlated with moisture content data obtained from the traditional gravimetric method. A predictive model was developed using an autoencoder and multilayer perceptron regression algorithms, achieving excellent performance with a coefficient of determination of 0.9832, root mean square error of 0.0103, and mean absolute error of 0.0088. This method exhibits non-destructive, non-contact, and rapid measurement capabilities, suitable for real-time and online applications, thereby contributing to the advancement of intelligent management in agriculture.