Abstract:Aiming at the problems of low automation and high cost of field spreading device for factory planting seedlings, a fully automated bilateral rail type field spreading device with a visual detection module for abnormal bumps in seedbed was designed. Firstly, the working principle of the tray spreading structure was analyzed, and then the structural design, force analysis and simulation analysis were carried out on the full-load operation condition of the tray spreading device. In order to prevent the abnormal bulge of seedbed from tilting the seedling tray when spreading the tray, which affected the survival rate of seedling refining, an abnormal bulge target recognition algorithm was proposed based on CBAM-YOLO v5n, and the improved YOLO v5n algorithm added the attention mechanism, and the average values of accuracy, recall, and average precision for the detection of the abnormal bulge target of the seedbed were respectively 98.1%, 91.7% and 94.9%, which were 1.2 percentage points, 1.7 percentage points and 0.9 percentage points higher than that of the original model, respectively. The developed tray-laying prototype was tested by orthogonal test method, and the test results showed that when the height of tray-laying was 90mm, the rotational speed of tray-laying mechanism was 550r/min, and the translational speed of tray-laying box was 0.14m/s, the highest tray-laying success rate was 96.4%, and after implanting the machine vision module, the tray-laying success rate can reach 99.3%. The designed tray spreading device can effectively reduce the labor intensity of manual tray spreading and reduce the labor cost of tray spreading.
