Locomotion

This is an Enhanced Lab Training Program, under Prof. Yue Wang’s Guidance.

In this project, we modeled a wheeled robot and implemented its motion control that responds to keyboard input in Gazebo. For navigation, We applies RRT* algorithm for global path planning and DWA algorithm for real-time local obstacle avoidance. Also, we implemented a SLAM system which uses laser odometry to collect data, and used the ICP algorithm for localization. Finally, the algorithms were successfully deployed and validated on a physical robot platform based on Raspberry Pi 4B.

Comparison of simulated and real-world algorithm performance. Left: simulated results showing planned (blue) vs. actual (green) path. Right: real-world deployment showing planned (pink) vs. actual (yellow) trajectory.