Overview
The Mobile Manipulation project applies fundamental principles from Modern Robotics by Kevin Lynch, including velocity kinematics, odometry, motion planning, trajectory generation, velocity control, and forward kinematics. The project focuses on planning and executing a trajectory for the end-effector of the youBot mobile manipulator, which consists of a mobile base with four mecanum wheels and a 5R robot arm.
The system performs odometry as the robot chassis moves, uses feedback control to guide the youBot to pick up a block from a specified location, transport it to a target area, and place it down.
Various controllers were tested, each with different levels of overshoot and convergence. Among them, the Feedforward PI controller proved to be the most effective for accurate trajectory tracking. The software architecture is built around three core components:
- Kinematics Simulator
- Trajectory Generation
- Feedforward Control
These components were seamlessly integrated within a control loop to enable precise motion and ensure the successful completion of the pick-and-place task.
View on GitHub