Development of an autonomous drone spraying control system based on the coefficient of variation of spray distribution

Pests and disease prevention has long been a key area of focus in precision agriculture research. While unmanned aerial spraying systems have advanced significantly and gained widespread adoption in recent years, challenges persist, including the high cost of precision spraying drones and issues related to uneven spraying and over- application with conventional systems. To address these limitations, this paper introduces a low-cost, versatile, and modular autonomous spraying control system that includes a ground base station and a spraying control assistant. The system integrates a spraying uniformity control algorithm based on a regression forest model, ensuring a coefficient of variation (CV) below 30 %. It also collects real-time environmental data to optimize the drone's spraying strategy. Environmental data and global positioning system's correction signals are transmitted from the ground base station to the onboard spraying control system (mobile station) via LoRa communication, enabling precise positioning and real-time adjustments during spraying. Indoor spraying simulation experiments demonstrate that the autonomous spraying control system achieved a CV within the standardized requirement in 15 out of 23 trials, with an overall predicted CV of less than 30 %. In outdoor experiments, using a hypothetical prescription map for targeted precision spraying, the system successfully completed all prescribed spraying zones. All targeted zones met directed spraying performance indicators exceeding 0.87, demonstrating high accuracy. The system shows significant potential for enhancing the precision spraying capabilities of conventional drones while reducing pest and disease control costs.