Design of Variable Spray System for Plant Protection UAV Based on CFD Simulation and Regression Analysis

Area of Science:

• Agricultural Engineering
• Precision Agriculture
• Robotics

Background:

• Multi-rotor unmanned aerial vehicles (UAVs) are crucial for plant protection in China.
• Non-uniform spray droplet distribution and drift reduce efficiency and harm the environment.

Purpose of the Study:

• To design, discuss, and verify a variable spray system for UAVs.
• To improve the uniformity and accuracy of agricultural spraying.
• To reduce environmental impact from pesticide application.

Main Methods:

• Computational fluid dynamics (CFD) simulations to analyze droplet deposition under various states.
• Training support vector regression and back propagation neural networks using simulation data.
• Multi-sensor fusion for real-time UAV offset and nozzle flow prediction.

Main Results:

• CFD simulation accuracy verified with wind velocity error < 1 m/s and deposition ratio error < 10%.
• An optimal regression model achieved a root mean square error of 6.5%.
• Farmland experiments demonstrated a deposition volume prediction error within 30%.

Conclusions:

• The developed variable spray system effectively improves spray deposition accuracy.
• The system offers a reliable solution for intelligent UAV spray applications.
• This research provides a valuable reference for advancing UAV-based agricultural technologies.