Vision-Based Hybrid Controller to Release a 4-DOF Parallel Robot from a Type II Singularity

Area of Science:

• Robotics
• Control Systems
• Human-Robot Interaction

Background:

• Parallel robots (PRs) offer high accuracy and dynamic performance for human-robot interaction.
• PRs face challenges like reduced workspace and type II singularities, which cause loss of control.
• Type II singularities require methods to maintain PR stiffness and control.

Purpose of the Study:

• To propose a novel hybrid controller for releasing a 4-DOF parallel robot from type II singularities.
• To utilize a real-time vision system for automatic configuration readjustment of the PR.
• To ensure patient safety during knee rehabilitation using the controlled robot.

Main Methods:

• Detection of type II singularities using the angle between two output twist screws (OTSs).
• Implementation of a hybrid controller integrating a real-time vision system.
• Automatic readjustment of PR configuration by actuating specific limbs identified via vision data.

Main Results:

• Successful release of a 4-DOF PR from type II singularity.
• Demonstration of smooth and controlled movements during singularity release.
• Validation of patient safety during the rehabilitation process.

Conclusions:

• The proposed hybrid controller effectively addresses type II singularities in parallel robots.
• Real-time vision systems can be integrated into control strategies for enhanced robot performance.
• This approach enhances safety and efficacy in robot-assisted knee rehabilitation.