On-line acquisition of fertilizer ingredient information based on multi-segment capacitive sensor combined with RLC resonance

• Published in: Flow measurement and instrumentation Vol. 106; p. 103029
• Main Authors: Li, Jianian, Ma, Yongzheng, Yang, Chuanke, Fang, Jiaoli, Liang, Jiawen
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2025
• Subjects: Capacitive sensors; Dielectric properties; Fertilizer detection; Frequency response; RLC resonance; Capacitive sensors; Frequency response; Dielectric properties; Fertilizer detection; RLC resonance
• ISSN: 0955-5986
• DOI: 10.1016/j.flowmeasinst.2025.103029

Irrigation-fertilization integration is a vital component of agricultural development, and the detection of fertilizer solution component information serves as a key link in this integration process. However, conventional detection methods for fertilizer solution components are plagued by deficiencies, including the incapacity to execute real-time online monitoring and a tendency to generate excessively large detection errors. In order to address these challenges, the study has designed a multi-segment capacitive sensor and proposed an online detection method based on RLC resonance and characteristic frequency response. The validity of the method is established through the utilization of urea, calcium superphosphate, potassium sulfate, potassium dihydrogen phosphate, ammonium dihydrogen phosphate, and compound fertilizers as test subjects. Firstly, the frequency response characteristics within the range of 1 kHz–30 MHz are investigated. The findings indicate that the sensor demonstrates optimal detection performance within the frequency range of 1 MHz–20 MHz Secondly, the characteristic frequencies of different fertilizer solutions are determined as 7.5 MHz, 9.5 MHz, 14.5 MHz, 3 MHz, 11.5 MHz, and 17.5 MHz based on the series resonance response characteristics. A concentration detection model based on the sensor's amplitude voltage is constructed, with determination coefficients (R2) all exceeding 0.9909. Finally, a detection strategy of multi-segment parallel operation with a “species-first-then-concentration" approach is adopted for qualitative and quantitative analysis of fertilizer solutions. Validation results show that the species recognition accuracy exceeds 94 %, and the concentration detection error is less than ±7.5 %. This research meets the demand for online detection of fertilizer solution components and concentrations in agricultural engineering, providing technical support for the development of irrigation-fertilization integration technology. [Display omitted] •A novel multi-segment capacitive sensor integrated with RLC resonance is developed for real-time, non-invasive detection of fertilizer components.•Characteristic frequencies of six fertilizer types are identified, enabling accurate classification with over 94% identification accuracy.•High-precision concentration detection models are built, with max detection error <±7.5% in 1%–10% concentration range.•The sensor shows excellent stability at 5°C–40°C and during prolonged operation, ensuring robust performance for agricultural use.