Navigation Research Of Warehouse Robot Based On SLAM

After entering the 21 st century,with the development of science and technology and the introduction of concepts such as Industry 4.0 and China Smart Manufacturing 2025,the traditional manufacturing industry has also undergone tremendous changes,with high labor costs,poor safety,low efficiency,and high error rates.The traditional warehousing and logistics industry has also followed the trend and moved towards intelligence.The navigation technology of warehousing robots is the key to promoting the intelligentization of the warehousing and logistics industry,and its development has attracted much attention.In this paper,for the warehouse robot system,the research on the SLAM-based warehouse robot navigation in the ROS environment is carried out,and the lidar is used as the main sensor to build an experimental platform to carry out related experiments in the simulation environment and the real environment.First,the research on the mapping part of the robot navigation system is carried out.The characteristics of maps in several two-dimensional environments are analyzed,and grid maps are selected as the representation of the navigation system.Aiming at the unreasonable weight distribution of particles and its dependence on the noisy motion model in the particle filter-based Gmapping algorithm,the method of feeding back the output information of the system to the motion model is adopted to reduce inefficient particle operations and effectively improve Gmapping.algorithm accuracy,and preprocess the map input into the navigation system to reduce map noise.Secondly,the research on the positioning part of the robot navigation system is carried out.For the traditional AMCL,the odometer with a large cumulative error is used as the basis for the auxiliary particle update,and the laser model is easily affected by the complex environment,which leads to the pose and the real output of the traditional AMCL.There is a problem of deviation in pose.An AMCL algorithm that introduces EKF and scan matching optimization is proposed,and the performance of the optimization algorithm is verified by experiments.Then,the research on path planning in the robot navigation system is carried out.In view of the large number of nodes in the expansion and maintenance list of the traditional A*path planning algorithm,and the lack of processing of redundant inflection points,combined with the working nature of the warehouse robot itself,the The traditional A* algorithm is improved and combined with the dynamic decision-making system designed in this paper to propose a global dynamic planning,which optimizes and improves the traditional A*algorithm in the three aspects of timeliness,smoothness and real-time dynamic processing capability of path planning,and simulates it in the Matlab environment Experiments verify the effectiveness.Finally,through the mobile robot experimental platform built,the simulation environment and real environment under the ROS system are used to experimentally analyze the robot’s mapping technology,positioning technology and path planning technology during the navigation process to verify the feasibility of the navigation system proposed in this paper.Experiments show that the system can well ensure that the warehouse robot can complete the navigation task in the warehouse environment.