Zhang Ruirui, Wen Yao, Yi Tongchuan, Chen Liping, Xu Gang. Development and application of aerial spray droplets deposition performance measurement system based on spectral analysis technology[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2017, 33(24): 80-87. DOI: 10.11975/j.issn.1002-6819.2017.24.011
    Citation: Zhang Ruirui, Wen Yao, Yi Tongchuan, Chen Liping, Xu Gang. Development and application of aerial spray droplets deposition performance measurement system based on spectral analysis technology[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2017, 33(24): 80-87. DOI: 10.11975/j.issn.1002-6819.2017.24.011

    Development and application of aerial spray droplets deposition performance measurement system based on spectral analysis technology

    • Abstract: To evaluate the droplet deposition in aerial spraying real-timely and accurately, aerial spray pattern measurement system was designed combining with spectral analysis and fluorescence excitation technology. The hardware of the system consisted of modules of information acquisition module, data acquisition module, and data processing module. FLAME-S-VIS-NIR micro spectrometer was selected as information acquisition module which is produced by Ocean Optics. Micro spectrometer was the core component of the aerial spray pattern measurement system. The acquisition module included microcontroller unit, droplet collection medium, Ultraviolet excitation light, stepper motors, and photoelectric limiter and so on. The software of the system includes the function of spectrometer connection, parameter setting, spectral data collection, display and storage. At first, the solution of fluorescent tracer with mass fraction of 0.5%, 1.0% and 1.3% was sprayed individually by the sprayer installed on the agricultural plant protection unmanned aerial vehicle. Droplet deposition was collected by droplet collection medium and water sensitive paper synchronously. The spectral characteristic curve of droplet collection medium was scanned and saved by the software of aerial spray pattern measurement system. The spectral characteristic curve of sample point was processed by savitzky-golay smoothing and standard normalized variate, the trend of spectral curve was analyzed. Without the effect of ultraviolet light on the band removal of 340-400 nm, the result which was observed and analyzed from the band range of 440-1 014 nm showed that the spectral band range of 450-460 nm presented a trough shape, and the spectral band range of 500-520 nm showed peak shape. Droplet deposition characteristic parameter which was obtained from the image analysis of water sensitive paper included impregnation area, area coverage and deposition. Compared with the result obtained by water sensitive paper, the analysis result indicated that the solution of fluorescent tracer on the droplet capture medium had produced significant fluorescence effect in the wavelength range of 450-460 nm and 500-520 nm. The spectral average value of the wavelength range of 450-460 nm and 500-520 nm was calculated. And the correlation coefficient of spectral average value and droplet deposition was up to 0.80. The results showed that it was feasible for the detection of droplet deposition characteristics based on spectral analysis and fluorescence excitation technique. The detection effect of droplet deposition with the different mass fraction of fluorescent tracer solution was analyzed. Compared with the mass fraction 0.5% and 1.3% of fluorescent tracer solution, the correlation coefficient between spectral average value and droplet deposition was more than 0.92 when the mass fraction of fluorescent tracer solution was 1.0%. Therefore, a fluorescent tracer solution with mass fraction 1.0% was adopted for the detection of performance test of the system. The performance test of the system was carried out in the field. Fifty nine sample points were collected effectively, and the multivariate linear regression model of the droplet deposition was built based on the spectral average value which was calculated by randomly selection of 40 sample points. The rest of 19 sample points was used to validate the multivariate linear regression model. The model decision coefficient was about 0.80, and the verification coefficient was about 0.83. The modeling accuracy can satisfy the requirements of droplet deposition characteristic parameter detection. This method could provide support for the detection of droplet deposition characteristics rapidly and continuously in aerial spraying.
    • loading

    Catalog

      /

      DownLoad:  Full-Size Img  PowerPoint
      Return
      Return