UR5e Pendulum Demo
This paper presents the development of a control law for a six-joint robot manipulator to balance an inverted pendulum using visual feedback. The proposed approach uses resolved motion control and the null space control techniques, which enable direct control of the robot manipulator's end-effector and filter out unwanted joint movements. While the control law is designed for the specific task of balancing an inverted pendulum, the key point is that resolved motion control is a highly precise technique that can be applied to a wide range of tasks. To evaluate the effectiveness of the approach, experiments were conducted in simulations to study the impact of visual feedback due to sampling rate. The paper concludes that task space control using resolved motion control can facilitate more natural human-robot collaboration and enable robots to perform complex tasks with high precision.
The study of a simple pendulum system can be traced back to Galileo's first experiments in the year 1602. From a control engineering perspective, the inverted pendulum model can be used in many engineering applications, ranging from the ambulation of bipeds on the ground to rockets or missiles in the sky [4], which makes it a good candidate to showcase the control of a robot arm with an emphasis in precision and speed. By convention, a fixed track is normally used as a physical constraint to ensure the 1-D horizontal movement of the cart. In addition to offering fine joint control, the robot arm used in this paper contains built-in sensors for precise joint state feedback.
TBA