Model and experiment of target-specific variable spraying based on canopy volume perception

To address the issues of pesticide deposition in non-target areas, pesticide drift leading to environmental pollution, and low pesticide utilization rates associated with traditional orchard spraying methods, a target-specific variable spraying prototype based on LiDAR (Light Detection and Ranging) canopy profiling was designed. The prototype employed a 3D LiDAR sensor as a detector and utilized the improved α-shape algorithm to extract the canopy contours, allowing for the accurate calculation of canopy volume. Based on the canopy volume, a pulse width modulation (PWM) pesticide dosage calculation model was established. The prototype integrates the offset of the droplets in the wind field to construct a spray-barrel pitch control model that fits the canopy contour. An orchard laser target-specific variable spraying control system and device were developed by incorporating a target-specific spraying control method. This device was mounted on a tracked power chassis to create an integrated target-specific variable-spraying machine prototype. Test results indicated that the droplet deposition density exceeded 80 $\text{droplets}\bullet {\text{cm}}^{‐2}$, achieving effective canopy coverage. In the designated test area, the pesticide solution usage for continuous spraying and target-specific variable spraying was 1.359 and 0.715 L, respectively, saving 47.96 % of pesticide with target-specific variable spraying. This system can achieve precise control of target-specific variable spraying based on changes in the position and volume of the tree canopies. The results of this study contribute to the rapid development and application of precision spraying technology in orchards.