An integrated free-space microwave transmission-based measuring system for non-destructive prediction sensing of moisture content and bulk density in wheat
Computers and Electronics in Agriculture, cilt.252, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 252
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.compag.2026.112147
- Dergi Adı: Computers and Electronics in Agriculture
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Applied Science & Technology Source, BIOSIS, Compendex, Environment Index, Geobase, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO), Technology Collection (ProQuest)
- Anahtar Kelimeler: Dielectric properties, Digital agriculture, Free-space microwave transmission, Moisture content, Wheat
- Sivas Cumhuriyet Üniversitesi Adresli: Evet
Özet
Moisture content and bulk density are critical parameters for wheat storage, handling, and post-harvest quality management. Measuring both rapidly and without damaging the grain remains a practical challenge. This study presents an integrated free-space microwave transmission-based measurement system for the simultaneous and non-destructive prediction of moisture content and bulk density in wheat. The system uses synchronized RF/LO operation, direct intermediate-frequency sampling, and Fast Fourier Transform-based vector processing to extract the attenuation and phase shift of microwave signals transmitted through wheat samples over the 2–4 GHz frequency range. From these transmission parameters, the dielectric constant (ε′), loss factor (ε″), and loss tangent (tanδ) were calculated and used, together with attenuation and phase shift, as model inputs. Four wheat cultivars widely grown in Türkiye, namely Bezostaya, İkizce, Kenanbey, and Tosunbey, were evaluated over four harvest years under varying moisture content and bulk density conditions, producing a dataset of 2,420 experimental trials. Thirteen regression algorithms were evaluated under single-frequency, reference-frequency, and multi-frequency scenarios. The results revealed a strong dependence of prediction performance on operating frequency. For moisture content prediction, the highest test performance was obtained at 3900 MHz using SVR-RBF (R2 = 0.9433, RMSE = 0.4505), followed by TabPFN-v2 (R2 = 0.9404, RMSE = 0.4621) and EBM (R2 = 0.9370, RMSE = 0.4751). For bulk density prediction, the best single-frequency performance was achieved at 3400 MHz with SVR-RBF (R2 = 0.8342, RMSE = 0.0137), closely followed by TabPFN-v2 (R2 = 0.8331, RMSE = 0.0138) and EBM (R2 = 0.8239, RMSE = 0.0141). These results demonstrate that the dielectric signatures captured by the developed free-space microwave transmission system contain sufficient information to predict moisture content and bulk density. The proposed system offers a compact, non-contact, and non-destructive sensing framework with potential for future sensor-level applications in post-harvest grain quality monitoring.