Guidance And Localization For Automatic Guided Vehicle Based On Contour Scanning

Automatic guided vehicle(AGV) is a kind of vehicle installed electromagnetic, magnetic, vision or laser equipment for guiding. It can drive along the specified path by itself and often used in transportation. With its advantages of intelligent, flexibility and cheapness, it has been widely used in logistics, warehousing and manufacturing. AGV improved the production efficiency and reduced costs in production.For AGV, the most important technology is localization and guidance of the vehicle. At present, there are some popular methods in the domestic market, such as magnetic guidance and laser guidance. Magnetic guidance needs to lay magnetic path on the surface of the ground, and laser guidance is required to place laser reflectors around the path which AGV is driving along. These two guidance methods are required to set marks for identification in advance. And visual guidance technology is still being studied, has not been massively applied in market and application. In order to solve the problem of the methods of guidance for AGV in the domestic market, this paper uses the laser radar to scan the environment and feedback to the AGV in the complex production environment. The research contents are summarized as followsTo process the data of contour scanning, ICP algorithm is used to process the data from laser radar, and obtain the position information of AGV. The ICP algorithm for point cloud matching in complex environment is analyzed and customized, so it can keep the speed of operation and the matching precision. The feasibility of the algorithm is verified by experiment.For localization, the position and posture data is obtained by using odometer for dead reckoning. The changes of position data obtained by contour scanning are used to correct the accumulative error of dead reckoning. The feedback error of contour scanning will be analyzed. The accumulated error generated in the contour scanning will be eliminated by pre storing the point cloud which scanned in the fixed position.For guidance, the kinematics model of AGV is established, and the process of trajectory tracking is analyzed. A kind of control law is used to control the AGV. Simulation experiments are conducted to analyze the function of the parameters in the control law, and to select the appropriate control parameters.The guidance method is tested in the actual production environment. The influence of the number and location of the pre stored point cloud, the position and posture error at the beginning of the path, and the velocity of AGV are analyzed. The feasibility of this method and the reliability of this method in the practical application are verified by experiments.