Design and Simulation of Intelligent Harvesting Device for Daylily

Abstract

In view of the prevalent issues of low efficiency, high labor demands, and elevated costs associated with daylily harvesting in most cultivating regions, an automated daylily harvesting device was meticulously designed. This device integrates advanced technologies to address the inherent challenges of daylily harvesting. Initially, SolidWorks was utilized to create a virtual prototype, and Ansys software was employed to simulate the picking action of the end effector, thereby validating the feasibility of its operation. Subsequently, the D-H method was adopted to model the robotic arm, and the Robotics Toolbox in MATLAB was leveraged to construct a digital representation of the robotic arm. This facilitated the analysis of its working interval and trajectory planning simulation. Furthermore, the Yolov7 model was incorporated to recognize daylily images, ensuring the accuracy of the picking process. The outcomes of this research demonstrate that the visual algorithm boasts a high recognition rate, and the chosen robotic arm and end-effector fully meet the design specifications, thereby effectively enhancing the automation level of daylily picking operations.