Design and field evaluation of a stratified variable spraying system in vineyards

During disease and pest control or nutrient solution spraying, grapevine’s morphology characterised the leaf curtain uneven distribution along the growing direction. Complex canopy structure and volume has a great impact on spray quality. This study proposes an air-assisted, stratified variable spraying system, which integrates a LiDAR sensing module with independently controlled multi-channel nozzles, and the spraying amount is regulated according to volumes of canopy divided by longitudinal layers. By combining adaptive DBSCAN clustering with an improved α-shape boundary extraction algorithm, the system segments canopy point clouds and reconstructs contours, thereby improving the accuracy of 3D volume estimation. A PWM duty cycle control model was introduced to dynamically adjust spray flow rates according to canopy volume and forward speed of spraying machine. The manual volume estimation and droplet coverage measurements are adopted in vineyard spraying experiments, it demonstrated that the system also performed well in adjusting spray volume in response to canopy structure. Compared with constant spraying, the variable spray approach of this study reduced pesticide use at different advancing velocity. Stratified variable spraying system displayed good operational stability, adapting flexibility to canopy structures while possessing high spray precision and efficacy. The study provide theoretical and technical support for variable spraying in complex crop scenarios.