Scholarlink

Purpose <p>This study explores the integration of spatial variation and pulse width modulation (PWM) technology in unmanned aerial vehicle (UAV)-based variable-rate pesticide application to enhance precision and sustainability in agriculture. This study aims to quantify the effects of spatial variation on spray distribution and develop predictive models for estimating spray amounts under varying mission conditions.</p> Methods <p>The study utilized a multi-rotor UAV equipped with a liquid spray system. The study involved eight UAV missions across a 200-cell field, with three distinct PWM settings (1300, 1450, and 1650) and two flight speeds (2.5&#xa0;m/s and 6.0&#xa0;m/s). Missions 1–6 focused on identifying key operational variables influencing spray amounts, while Missions 7–8 validated the predictive models. Random forest and pruned decision tree models were used to analyze factors influencing “Flight Time per Cell (seconds)” and “Spray Amount (L).”</p> Results <p>Spatial variation significantly impacted spray distribution across the UAV missions. The study identified “Number of Cells,” “UAV start position,” and “Velocity” as key factors affecting spray distribution. The machine learning models accurately predicted spray amounts, with discrepancies noted between aerial and ground spray rates, especially at higher PWM settings.</p> Conclusions <p>This study underscores the importance of integrating real-time data and predictive modeling for UAV-based pesticide application. By addressing spatial variation and operational complexities, the findings contribute to improving spray efficiency, reducing chemical wastage, and promoting environmentally sustainable agricultural practices. Future research should incorporate diverse environmental conditions and intermediate flight speeds to further refine predictive models and enhance their real-world applicability.</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

A Study of Spray Volume Prediction Techniques for Variable Rate Pesticide Application using Unmanned Aerial Vehicles

  • 期刊论文

  • 发表日期：

    2025年1月7日

查看全文
  • Dasom Jeon,
  • Hyun-jin Jung,
  • Kyung-do Lee,
  • JungGon Han,
  • Chanwon Park,
  • Seunghoon Han,
  • Hojin Kim

摘要

Purpose

This study explores the integration of spatial variation and pulse width modulation (PWM) technology in unmanned aerial vehicle (UAV)-based variable-rate pesticide application to enhance precision and sustainability in agriculture. This study aims to quantify the effects of spatial variation on spray distribution and develop predictive models for estimating spray amounts under varying mission conditions.

Methods

The study utilized a multi-rotor UAV equipped with a liquid spray system. The study involved eight UAV missions across a 200-cell field, with three distinct PWM settings (1300, 1450, and 1650) and two flight speeds (2.5 m/s and 6.0 m/s). Missions 1–6 focused on identifying key operational variables influencing spray amounts, while Missions 7–8 validated the predictive models. Random forest and pruned decision tree models were used to analyze factors influencing “Flight Time per Cell (seconds)” and “Spray Amount (L).”

Results

Spatial variation significantly impacted spray distribution across the UAV missions. The study identified “Number of Cells,” “UAV start position,” and “Velocity” as key factors affecting spray distribution. The machine learning models accurately predicted spray amounts, with discrepancies noted between aerial and ground spray rates, especially at higher PWM settings.

Conclusions

This study underscores the importance of integrating real-time data and predictive modeling for UAV-based pesticide application. By addressing spatial variation and operational complexities, the findings contribute to improving spray efficiency, reducing chemical wastage, and promoting environmentally sustainable agricultural practices. Future research should incorporate diverse environmental conditions and intermediate flight speeds to further refine predictive models and enhance their real-world applicability.