Abstract
This project has conducted research on robot path planning based on VisualSLAM. The main work of this project is as follows: (1) Construction of VisualSLAM system. Research has been conducted on the basic architecture of VisualSLAM. A Visual SLAM system is developed based on ORB-SLAM3 system, which canconduct dense point cloud mapping. (2) The map suitable for two-dimensionalpath planning is obtained through map conversion. This part converts the densepoint cloud map obtained by Visual SLAM system into an octomap and thenperforms projection transformation to the grid map. The map conversion convertsthe dense point cloud map containing a large amount of redundant mapinformation into an extremely lightweight grid map suitable for path planning.(3) Research on path planning algorithm based on reinforcement learning. Thisproject has conducted experimental comparisons between the Q-learningalgorithm, the DQN algorithm, and the SARSA algorithm, and found that DQN isthe algorithm with the fastest convergence and best performance inhigh-dimensional complex environments. This project has conducted experimentalverification of the Visual SLAM system in a simulation environment. Theexperimental results obtained based on open-source dataset and self-madedataset prove the feasibility and effectiveness of the designed Visual SLAMsystem. At the same time, this project has also conducted comparativeexperiments on the three reinforcement learning algorithms under the sameexperimental condition to obtain the optimal algorithm under the experimentalcondition.