Research on ore sorting manipulator based on improved YOLOv7 algorithm

Abstract

In order to solve the problem of scrap iron, large ore and large tree roots that are difficult to be crushed by the crusher on the ore sorting conveyor belt in the mine, an ore sorting manipulator was proposed to sort the foreign matter on the conveyor belt. Firstly, the attention mechanism CA is introduced into the feature layer extracted by the backbone network of YOLOv7 algorithm to distinguish and locate the ores and foreign bodies on the conveyor belt, which improves the detection accuracy of the algorithm. The Map and F1 values of ores, large ores, scrap iron blocks and large roots are compared and analyzed. The Map values of four of them are good, close to 90%. Then, Matlab was used to simulate and analyze the trajectory planning of the robotic arm, and it was concluded that the optimization result of the robotic arm using the adaptive particle swarm optimization algorithm was 25.6, which was better than that of the robotic arm using the standard particle swarm optimization algorithm (52.7), and the time was also optimized by 50%.