Raven Control

Problems:

  1. Raven is a cable driven power transmission robot. The use of flexible cable based power transmission often causes relative motion between the motor actuator and mechanism joint during operation due to the elasticity of the cable.  Moreover friction and forward kinematic uncertainties are also causing the actual position of the end-effector to deviate from the expected position.
  2. Gravitational loads on robotic manipulators often present a large, nonlinear contributions to static and dynamic control. In many cases, this load can be calculated and compensated for when the positions of the robot links are known with respect to a base frame. However, this base frame is stationary with respect to gravity and it will not work when the base frame is not stationary.

Approach:

We propose to overcome these two issues by (1) better state estimates and (2) PD + dynamic gravity compensation controller or PD + DG for short.   State space control can be used to improve performance but may require state estimates. For nonlinear systems the Unscented Kalman Filter (UKF) provides an efficient way to obtain state estimates.   In this project, the researchers in BioRobotics Lab will be first investigating the application of UKF on the Raven, and second integrating UKF on the Raven with PD + DG controller.

In the above video, Andy Lewis demonstrates the effectiveness of gravity compensation on Shoulder joint (joint 0) on both arms.

 

Team:

BioRobotics Lab: Mohammad Haghighipanah, Yangming Li, Blake Hannaford

Alumni: Andrew Lewis, Sina Nia Kosari, Hawkeye King

Publications:

Srikrishnan Ramadurai, SN Kosari, H. Hawkeye King, HJ Chizeck, Blake Hannaford,, ‘Application of Unscented Kalman Filter to a Cable Driven Surgical Robot: A Simulation Study,,’ 2012 IEEE International Conference on Robotics and Automation, St. Paul-Minneapolis,, May 2012..

Andrew Lewis, ‘Dynamically Evaluated Gravity Compensation for the RAVEN Surgical Robot,’Department of Mechanical Engineering, August 2013.