Research On SLAM And Path Planning Of Indoor AGV

As an important part of mobile robot,Automated Guided Vehicle(AGV)plays an important role in various intelligent fields.In AGV autonomous navigation system,Simultaneous Localization and Mapping(SLAM)and path planning are the core technical problems to be solved.In this dissertation,aiming at the indoor AGV autonomous navigation technology,an indoor AGV working platform based on Robot Operating System(ROS)is built,and the slam and path planning problems are systematically studied combined with laser technology.Finally,the algorithm is improved and its performance is verified.The main work of this dissertation is as follows: in terms of location and mapping,aiming at the problems of rough mapping contour,low positioning accuracy and lack of diversity of particles in the traditional Gmapping algorithm,an improved Gmapping algorithm based on Region Seeds Growing(RSG)is proposed.The algorithm combines the region growing method to extract the line features of point cloud data,and completes the point cloud registration through the position and angle of line features,The Extended Kalman Filter(EKF)is used to fuse the information of odometer and Inertial Measurement Unit(IMU).At the same time,in order to maintain the diversity of particles,the Elite Genetic theory is introduced into the resampling process of Gmapping.Finally,by comparing the positioning and mapping results of the original algorithm and the improved algorithm in the same environment,it is proved that the improved algorithm in this dissertation has more accurate positioning advantages of clearer mapping;In terms of path planning,in order to solve the shortcomings of traditional Jump Point Search(JPS)algorithm in path planning,such as low search efficiency and uneven path planning,this dissertation proposes an improved path planning algorithm based on heuristic two-way search.The algorithm introduces the two-way search mechanism into JPS algorithm and integrates the improved JPS algorithm and Dynamic Window Approach(DWA)algorithm,it can reduce the path search time and avoid obstacles.The improved JPS algorithm proposed in this dissertation is compared with the existing common algorithms in the simulation environment.The results show that the improved JPS algorithm proposed in this dissertation effectively reduces the path search time and search nodes,and the combined path planning algorithm of improved JPS and DWA can avoid obstacles on the basis of global optimization.Finally,an indoor AGV navigation experimental platform is built based on ROS,and slam mapping and path planning experiments are carried out in the actual indoor scene to complete the evaluation of the improved algorithm.The results show that compared with the original algorithm,the improved algorithm proposed in this dissertation shows better practicability,which is embodied in: the improved Gmapping algorithm has higher accuracy and clearer drawing outline;The improved JPS algorithm has faster search speed and smoother planned path;Combined path planning based on improved JPS can realize dynamic obstacle avoidance on the basis of global path optimization.Figure [56] Table [9] Reference [82]...