Sebastian Illi
Efficient Microcontroller-Operated Motion Control for Resource-Constrained Mobile Robots
Abstract
This thesis focuses on designing and implementing a motor control interface for small
mobile robots with a differential drive motor setup, with particular emphasis on
developing a closed-loop Proportional-Integral-Derivative (PID) controller. A detailed
overview of the PID control method and its integration on low-level signal processors
is provided. The primary aim is to offload computational tasks related to closedloop
control from the main board computer to a dedicated signal processor, enabling
the creation of a differential drive motor controller. The work further explores the
communication between the signal processor and the Robot Operating System 2 (ROS 2)
framework, including implementing action servers for simplified motion control.
Additionally, the relative odometry of the robot is accurately calculated, ensuring
precise tracking of the robot’s position and orientation during operation.
Experimental results show optimal PID parameters for different motor setups. While
the robot demonstrates slight inaccuracies during initial movement, its linear motion
tolerance remains within 1 cm, with a control loop reaction time of 10 ms.
The effectiveness of the implementation is evaluated through a maze-solving scenario,
comparing a simplified navigation approach with the ROS 2 Nav2 localization software.
This work demonstrates the potential of using signal processors for motor control in
mobile robots and highlights areas for future improvement in sensor integration and
localization capabilities.