Grabbing Strategy of Single-Line Lidar Imaging of Stack Based on Genetic Algorithm

Today, single-line lidar is widely used in automated production and autonomous driving, object recognition and imaging is a key part of the field of lidar research. With the growing demand for radar collaborative robots, the use of radar imaging to guide robots to improve work efficiency is becoming a key research for intelligent devices. In this paper, the application of lidar imaging in the brewing robot grasping material pile is mainly described. Among them, we propose a dynamic calculation volume and material handling strategy algorithm based on fermentation heap point cloud data. In order to calculate the volume of the pile, we pre-process the image twice, and through two corrections and de-distortion, we can accurately calculate the volume of the pile. For the material pile grasping algorithm, we propose a material cutting and grasping method based on genetic algorithm. In contrast to traditional manual target calibration and prior fixed-point capture, our method is designed for fully automatic target calibration. As the mobile robot travels along the channel toward the target location, the lidar scans the material stack and combines the measurement data and motion estimation information collected at successive times to form an initial map. Then, the initial graph robustness optimization and data processing are calculated to calculate the next step of the robot to grab the point information. Simulation and experimental results show the efficiency and accuracy of the proposed method.