Date of Award
August 2024
Degree Type
Dissertation
Degree Name
Doctor of Philosophy
Department
Engineering
First Advisor
Mohammad H Rahman
Committee Members
Amol D Mali, Ilya V Avdeev, Nathaniel E Stern, Md Rasedul Islam
Keywords
Control, CoppeliaSim, REAL-TIME, Robotics, Trajectory, TWINCAT 3
Abstract
The field of robotics faces a significant challenge in bringing together innovative research and practical, industry-ready solutions. This divide is primarily due to the high cost of industrial robots, reliance on closed, proprietary software, focus on reliability over innovation, and the complexity of real-time control systems. Additionally, the fragmented landscape of communication protocols across different vendors complicates the integration of diverse components, often leading researchers to depend heavily on simulations that may not translate effectively to real-world applications.This dissertation addresses these challenges by introducing a novel, unified framework for real-time robot control and hardware-in-the-loop (HIL) simulation within the industry adopted TwinCAT 3 environment. This framework enhances affordability and accessibility by enabling direct control of various robotic hardware, including a commercial UFactory xArm robot, using open communication protocols and eliminating the reliance on expensive proprietary control systems. By leveraging TwinCAT 3’s real-time capabilities and support for industrial communication standards like EtherCAT, Modbus RTU, and Modbus TCP, the framework offers a deterministic platform for controlling a heterogeneous set of robots, including a four-mecanum-wheeled mobile robot and a custom-built robotic manipulator. Key innovations of this research include the development of a cost-effective and accessible platform for real-time robot control, facilitated by the industrially relevant TwinCAT 3 software and open communication protocols. The framework’s tight integration with MATLAB/Simulink allows for rapid control prototyping and HIL simulation, addressing the gap between simulation and real-world implementation. This research focuses on designing and validating a scalable Mecanum wheel mobile robot platform, integrating both a custom-built six-degree-of-freedom (6-DOF) robotic arm and a commercially available desktop collaborative robot. A generalized dynamic model for Mecanum wheel mobile robots and a 9-DOF mobile manipulator system is proposed, employing the Denavit-Hartenberg (DH) convention and the iterative Newton-Euler formulation. Trajectory generation algorithms, including jerk-limited trajectories and B-spline curves, are investigated for optimal motion time and synchronized movement across multiple degrees of freedom. The research validates multiple control strategies through extensive simulations in MATLAB/Simulink and CoppeliaSim, demonstrating their effectiveness in precise trajectory tracking and stable motion control in both simulation and real-world experiments. This dissertation significantly contributes to the field by developing a novel, open-architecture real-time motion control platform. The platform's modular architecture and the use of easy to access software and hardware components enhance its adaptability for diverse robotic systems, paving the way for future research and development in robotics and automation. This solution provides a valuable alternative to traditional, vendor-locked control systems, enabling researchers and engineers to create more flexible, scalable, and efficient robotic systems.
Recommended Citation
Assad Uz Zaman, Md, "REAL-TIME CONTROL OF ROBOTIC SYSTEMS: A SCALABLE PLATFORM WITH TWINCAT 3" (2024). Theses and Dissertations. 3549.
https://dc.uwm.edu/etd/3549