The growing interest in employing industrial robots for tasks like machining, joining, painting, assembly, and more, stems from their potential to improve workflow and efficiency. However, the basic mechanical structure of these robot manipulators often leads to low rigidity and dynamic issues, negatively impacting their overall accuracy and consistency. For specific tasks such as spot welding or pick-and-place operations, which require precise positioning, certain offline teaching or compensation techniques are available. When it comes to more complex tasks like robotic milling, the robots’ high structural compliance and the presence of strong, repetitive forces can cause significant vibrations and sometimes chatter. Moreover, the position-dependent characteristics of these robots bring about numerous frequency response functions (FRFs), adversely affecting their ability to accurately follow a set trajectory.
Our research is focused on developing algorithms and strategies to address these challenges. We are exploring the use of external sensors in conjunction with closed-loop offline and online compensation methods, and conducting experiments to validate these algorithms. One key area of our study involves employing data-driven techniques to precisely determine mode-dependent modal parameters across different workspace areas. We aim to optimize robot poses through static and dynamic stiffness models, ensuring closed-loop stability.
Our work also includes the creation of hybrid models that combine mathematical models with sensory data. This approach is particularly effective in online chatter detection during complex robotic milling operations, and it enhances the precision of the robot’s end-effector. Additionally, we are investigating data-driven and model-based methods to improve gain scheduling techniques. These methods adjust controller gains based on the joint configuration along a set trajectory. Other aspects of our research involve trajectory planning with real-time position error feedback control and offline path compensation. These techniques, which utilize external sensors and physics-based models, aim to increase the machining accuracy of compliant parts.
Publications
Cvitanic T, Melkote SN. A new method for closed-loop stability prediction in industrial robots. Robotics and Computer-Integrated Manufacturing. 2022 Feb 1;73:102218.
Nguyen V, Melkote S. Hybrid statistical modelling of the frequency response function of industrial robots. Robotics and Computer-Integrated Manufacturing. 2021 Aug 1;70:102134.
Lanzillotta, L., Berenji, K.R., Cvitanic, T., Brown, A. M., Freeman, P. L., Melkote, S. N., Simulation-based Optimal Gain Scheduling of Industrial Robotic Manipulators, Robotics and Computer-Integrated Manufacturing
Contact information
- Kave Rahimzadeh Berenji: Kaveh.berenji@gatech.edu
- Mitchell Henderlong: mhenderlong3@gatech.edu