Steven A Velinsky

Singularity Avoidance for the 3-RRR Mechanism Using Kinematic Redundancy

Based on its simple structure, base-fixed actuators, high payload capacity, high accuracy, and high mechanical rigidity, the 3-RlowbarRR mechanism is a valuable planar parallel manipulator. However, the 3-RlowbarRR mechanism is known to have singular loci within its workspace affecting its use. In this paper, singularity avoidance of the 3-RlowbarRR mechanism using kinematic redundancy is presented. First, singularity analysis of the proposed 3-RlowbarPlowbarRR mechanism is described and a simple and effective redundancy resolution algorithm based on local optimization suitable for real-time control is developed. Here, the cost function in the optimization is designed to avoid the most problematic singularity configurations, where the end-effector can be locally moved even though all actuated joints are locked. Results from simulation show that the resultant 3-RlowbarPlowbarRR mechanism can be used to avoid singularities associated with the 3-RlowbarRR mechanism, and enlarges the usable workspace.

Human-Robot Collision Detection and Identification Based on Wrist and Base Force/Torque Sensors

In this paper, a collision detection and identification method of a manipulator, using base and wrist force/torque sensors, is presented. An impact model is used to simulate the interaction between the manipulator and the human or environment. A neural network approach and a model based method are developed to detect the collision forces and disturbance torques on the joints of the manipulator. The experimental results illustrate the validity of the developed collision detection and identification scheme.