| In modern industrial production,equipment such as forklifts and trailers requires a large number of people to participate,and the cost is high and the transportation efficiency is low.It has become increasingly difficult to meet the production needs of modern industry.As an automatic transport vehicle,AGV can be widely deployed and has a wide range of motion.It satisfies the various needs of the factory.Modern factories have also begun to use AGV to transport materials.In industrial production,the accuracy requirements for AGV’s autonomous positioning are also getting higher and higher.It is of great significance to study the positioning algorithm of AGV to meet the positioning requirements of AGV map update and path planning.The main research contents of this article are as followsAiming at the problems of low calculation efficiency and poor real-time performance of the AGV positioning algorithm,the laser SLAM algorithm was selected to draw the grid map,and the world coordinate system,the car coordinate system and the laser coordinate system were established around the AGV.By analyzing the odometer and its error sources,Establish an odometer motion model,and then establish an observation model for the LMS111-10100 Lidar selected in this experiment to provide support for the following positioning algorithm,use the laser sensor of the research group to obtain data,perform a grid map simulation to create an experiment,and Performing binarization and natural localization processing reduces the calculation amount of the algorithm and meets the real-time requirements of positioning.Aiming at the AGV positioning problem,the AMCL algorithm is selected as the basic positioning algorithm of the car,and the laser model positioning simulation is performed with matlab.In the simulation,the particle set is completed after 2 displacement times,and the positioning error is about 3cm.In the experiment,it is found that the positioning algorithm has a problem of particle set dispersion.It may be due to the fact that the cumulative error of the odometer cannot be eliminated.As a result,the average weight of the particle set is reduced.During resampling,the algorithm will not trust the particle set.Poses,adding random particles to the global map,resulting in the problem of particle set dispersion,so an AMCL positioning algorithm based on the fusion of EKF odometer and inertial component data is proposed.By using EKF to odometer and inertial component data Fusion to establish the robot motion model,thereby correcting the problem of odometer error,increasing positioning accuracy and algorithm stability.Aiming at the problems of low positioning accuracy and scattered points in the positioning algorithm,experiments were conducted in the laboratory environment with an independently designed experimental platform.Through the comparative analysis of the positioning accuracy of the global positioning in the environment,the results show that the improved algorithm’s positioning accuracy is guaranteed to Within 2.5cm,the angular positioning accuracy is doubled compared to the original algorithm.In the experiment of robot kidnapping,the algorithm relocated the car after about 13s,and the positioning accuracy remained within 1 grid size,which verified the positioning failure recovery ability of the algorithm. |
