Project Details
Projekt Print View

Fault diagnosis and tolerance for elastic actuation systems in robotics: physical human-robot interaction

Subject Area Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
Human Factors, Ergonomics, Human-Machine Systems
Term from 2015 to 2022
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 277880821
 
Final Report Year 2024

Final Report Abstract

Ensuring safety and reliability is important for robotic systems, in particular for wearable robots such as prostheses and orthoses, which exhibit a tight human-robot interaction. Introducing elastic elements between the robotic system and the user is a promising approach to improve safety by reducing impact forces through elastic deformation. Yet, attaining reliable motion control can be challenging, as it might require accurate knowledge of the actuator properties. This project focused on improving the reliability of elastic actuators in physical human-robot interaction (pHRI). Faults that alter the elastic properties of such actuators can lead to a loss of precision or unsafe behavior. In the project, we developed a general fault-tolerant control strategy that can detect and compensate for elastic faults in real-time, to ensure the correct operation and safe interaction. Functional experiments were conducted to evaluate impedance control of elastic actuators in combination with fault detection methods. Through parameter adaptation, the control strategy is capable of accurate motion control under emulated fault conditions and user interaction. The general approach is applicable to elastic actuators with different mechanical designs, considering nonlinear and redundant characteristics, and ensures fault-tolerant interaction stiffness. Psychophysical and psychometric studies were conducted to evaluate how users experience and react to elastic faults during interaction. The first experiment investigated the thresholds of human stiffness perception and showcased the importance of virtual inertia shaping for interaction stiffness adaptation. A second user study with a knee orthosis examined the impact of elastic faults during walking with torque support at the knee joint. Results show that participants were able to clearly perceive changes in support level due to faults. Yet, when said faults were compensated, the differences in support level perceived by the participants were lower. Overall, the results of the project demonstrate the effectiveness of fault diagnosis and tolerance techniques for elastic actuation systems in providing accurate trajectory tracking and supporting human, e.g., during walking. The project highlights the importance of developing techniques that can detect and mitigate faults in real-time, ensuring that the robotic systems remain safe and reliable during operation in pHRI applications.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung