Conjugate-Gradient Method for Propulsion Optimization and Flight Stability of an Agricultural Drone Sprayer

Abstract

Conventional pesticide and liquid fertilizer spraying methods in agriculture remain inefficient because they require intensive labor, long operational time, inconsistent spraying quality, and direct human exposure to agrochemicals. Improving the flight performance of agricultural drone sprayers is therefore essential to achieve accurate, stable, and efficient spraying operations under varying payload conditions. This study presents the design, optimization, realization, and testing of an agricultural drone sprayer for pesticide and liquid fertilizer applications in precision agriculture. The study focused on optimizing the drone's overall flight performance using the conjugate-gradient method. The lifting power analysis considered several parameters, including payload mass, thrust generation, battery efficiency, and aerodynamic resistance. The optimization process was performed iteratively to minimize thrust errors and improve flight stability during low-altitude spraying operations. The simulation results showed that the optimization error decreased from 12.5% to 0.22% after fifteen iterations, while the flight stability index improved from 62% to 98%. These results indicate that the conjugate-gradient method effectively optimized the propulsion configuration, thereby enhancing the drone hovering stability and maneuverability. The developed prototype integrated a quadcopter propulsion system, spraying mechanism, GNSS/GPS-based autonomous navigation, and real-time telemetry monitoring. Experimental testing demonstrated stable flight performance, efficient spraying capability, and reliable operation under varying payload conditions. This study advances agricultural drone optimization by demonstrating the effectiveness of the conjugate-gradient method in improving flight performance, providing a scientific basis for developing more reliable and efficient unmanned aerial systems for precision agriculture.