Virtual Forward Dynamics Models for Cartesian Robot Control

Abstract

In industrial context, admittance control represents an important scheme in programming robots for interaction tasks with their environments. Those robots usually implement high-gain disturbance rejection on joint-level and hide direct access to the actuators behind velocity or position controlled interfaces. Using wrist force-torque sensors to add compliance to these systems, force-resolved control laws must map the control signals from Cartesian space to joint motion. Although forward dynamics algorithms would perfectly fit to that task description, their application to Cartesian robot control is not well researched. This paper proposes a general concept of virtual forward dynamics models for Cartesian robot control and investigates how the forward mapping behaves in comparison to well-established alternatives. Through decreasing the virtual system's link masses in comparison to the end effector, the virtual system becomes linear in the operational space dynamics. Experiments focus on stability and manipulability, particularly in singular configurations. Our results show that through this trick, forward dynamics can combine both benefits of the Jacobian inverse and the Jacobian transpose and, in this regard, outperforms the Damped Least Squares method.

A practical DMPs Implementation for Skill Creation and Teleoperation with Assistive Manipulators

“…H is a dynamics-shaped joint space inertia matrix that leads to better linearization and convergence of the IK solver. We refer the interested reader to [17] for more details on this aspect. Since Eq.…”

Optimizing Non-diagonal Stiffness Matrix of Compliance Control for Robotic Assembly Using Deep Reinforcement Learning

“…For some industrial robots that only have a position control interface, they only need to attach a six-dimensional force sensor on the end effector to achieve compliant control. The Virtual Forward Dynamics Model [3] is to concentrate the mass on the end effector by modifying the mass ratio of the end effector to the rest of the axes. This approach makes the robot's dynamics in cartesian space become linear and have better stability in the robot's singular position.…”

“…Hogan [2] originally proposed the concept of impedance control to describe the relationship between external force and displacement through a mass-spring-damper system to solve the stable transition from unconstrained to constrained motion. Reference [3] proposed a new compliant control method from the perspective of robot dynamics. This method creates the virtual forward dynamics model by reducing the link mass of the other axes relative to the end effector.…”

Contact Skill Imitation Learning for Robot-Independent Assembly Programming

“…During deployment, the LSTM-MDN model predicts 6dimensional Gaussian distributions, from which we sample force-torque strategies for robot control in each time step. We use the forward dynamics-based control approach of our previous work [41] to realize Cartesian force control on the robotic manipulator. Its implementation is available opensource 1 .…”

“…in which H is a linearized joint space inertia matrix that is computed in each control cycle. We refer to our work [41] for more details on this aspect. We then obtain the joint positions q with the Euler forward method according to…”