Experimental investigation on adaptive robust controller designs applied to constrained manipulators

Area of Science:

• Robotics
• Control Systems Engineering
• Sensor Technology

Background:

• Existing static force/moment sensors limit dynamic robotic applications.
• Robotic manipulators often face model uncertainties and external disturbances.
• Adaptive nonlinear control is crucial for robust robotic performance.

Purpose of the Study:

• To design and construct a dynamic 3D force/moment sensor as an alternative to static load cells.
• To experimentally investigate the performance of adaptive nonlinear H∞ control methods for constrained manipulators.
• To evaluate coordinated position and force control under model uncertainties and disturbances.

Main Methods:

• Development of a dynamic 3D force/moment sensor for robotic end-effectors.
• Implementation of four adaptive nonlinear H∞ control strategies using neural networks and fuzzy systems.
• Two modeling strategies: estimating uncertainties with a known nominal model, and estimating dynamics with an unknown model.
• Experimental validation using a planar manipulator and the developed dynamic force sensor.

Main Results:

• The dynamic 3D force/moment sensor provides a viable feedback signal for robotic end-effectors.
• Adaptive control methods demonstrated effectiveness in managing uncertainties and disturbances.
• Comparative analysis highlighted the performance differences between the two modeling strategies.

Conclusions:

• The developed dynamic sensor offers an advanced solution for robotic force feedback.
• Adaptive nonlinear H∞ control, particularly with intelligent systems, enhances manipulator performance in uncertain environments.
• The study provides valuable insights into sensor design and control strategies for advanced robotics.