Variable spraying technology for agricultural drones is key to achieving precision pesticide application. However, the response delay of the spraying system can cause offset in the deposition position of the liquid, affecting operational accuracy. This paper investigates the dynamic response characteristics of brushless motor-driven variable spraying systems, analyzing the causes of response lag and its impact on spraying precision. A robust control strategy based on an Extended State Observer (ESO) is proposed to optimize the motor's dynamic response. Research indicates that by establishing a motor mathematical model including disturbance compensation and employing a cascaded ESO to estimate and suppress system disturbances, the motor speed tracking performance can be significantly enhanced. Combined with the mathematical mapping relationship between PWM duty cycle and spray flow rate, the optimized control system effectively shortens response time and reduces spraying position errors, providing a theoretical basis for the precision quantification of variable spraying operations by agricultural drones.