KUKA LWR-IV Research Articles

On passivity-based trajectory tracking for robotic manipulators combining PD+ and Slotine-Li control

Abstract In this paper, a tracking controller for robots is analyzed, which allows both the specification of the convergence rate of the tracking errors and the parameterization of a desired contact stiffness and damping. It is shown that the control approach can be interpreted as a generalization of the well-known PD+ and Slotine-Li controllers, combining the benefits of both approaches. Although mentioned in the literature before, no thorough theoretical and practical analysis of the aforementioned passivity-based control concept has been performed so far. In this work, the implications of the gains are discussed w.r.t. convergence and interaction properties, addressing possible tuning strategies. Finally, the performance of the controller is evaluated in terms of tracking and interaction experiments on a KUKA LWR IV+.

Adaptive Tracking Control With Uncertainty-Aware and State-Dependent Feedback Action Blending for Robot Manipulators

Adaptive control can significantly improve tracking performance of robot manipulators subject to modeling errors in dynamics. In this letter, we propose a new framework combining the composite adaptive controller using a natural adaptation law and an extension of the adaptive variance algorithm (AVA) for controller blending. The proposed approach not only automatically adjusts the feedback action to reduce the risk of violating actuator constraints but also anticipates substantial modeling errors by means of an uncertainty measure, thus preventing severe performance deterioration. A formal stability analysis of the closed-loop system is conducted. The control scheme is experimentally validated and directly compared with baseline methods on a torque-controlled KUKA LWR IV+.

Enabling impedance-based physical human–multi–robot collaboration: Experiments with four torque-controlled manipulators

Robotics research into multi-robot systems so far has concentrated on implementing intelligent swarm behavior and contact-less human interaction. Studies of haptic or physical human-robot interaction, by contrast, have primarily focused on the assistance offered by a single robot. Consequently, our understanding of the physical interaction and the implicit communication through contact forces between a human and a team of multiple collaborative robots is limited. We here introduce the term Physical Human Multi-Robot Collaboration (PHMRC) to describe this more complex situation, which we consider highly relevant in future service robotics. The scenario discussed in this article covers multiple manipulators in close proximity and coupled through physical contacts. We represent this set of robots as fingers of an up-scaled agile robot hand. This perspective enables us to employ model-based grasping theory to deal with multi-contact situations. Our torque-control approach integrates dexterous multi-manipulator grasping skills, optimization of contact forces, compensation of object dynamics, and advanced impedance regulation into a coherent compliant control scheme. For this to achieve, we contribute fundamental theoretical improvements. Finally, experiments with up to four collaborative KUKA LWR IV+ manipulators performed both in simulation and real world validate the model-based control approach. As a side effect, we notice that our multi-manipulator control framework applies identically to multi-legged systems, and we execute it also on the quadruped ANYmal subject to non-coplanar contacts and human interaction.

Symplectic Discrete-Time Control of Flexible-Joint Robots: Experiments with Two Links

Abstract We present the application and experimental validation of discrete-time control implementation based on symplectic integration - in short referred to as symplectic discrete-time control - to a serial robot manipulator with flexible-joints. Applying the implicit midpoint rule as a simple and popular symplectic integration scheme to both the model of the sampled system and the desired target dynamics, yields a discrete-time control law with the same structure as in continuous time. The arguments are, however, predicted stage values on the subsequent sampling interval resulting from a half implicit Euler step for the target system. The approach can be easily extended to a purely position-based feedback law. The experimental validation on a lightweight robot (KUKA LWR IV+) confirms the expected improvements at low sampling rates, with an increase in the assignable closed-loop stiffness of up to 29 %.

Swing-Up of a Spherical Pendulum on a 7-Axis Industrial Robot

This work demonstrates the swing-up of a custom-built spherical pendulum, which is mounted on the 7-axis industrial robot Kuka LWRIV+. The swing-up trajectory for this system with nine degrees of freedom is calculated offline in a three-step process: First, an optimal control problem for the pendulum model with fewer states and inputs is solved. Then, using the inverse kinematics of the robot model and the forward dynamics of the complete system, this trajectory is converted to a feasible starting solution for the third step, which is solving the optimal control problem for the complete system. Finally, the swing-up is shown experimentally using a two-degrees-of-freedom control structure comprising the calculated trajectory for the feedforward control and a time-variant linear quadratic regulator (LQR) as a feedback controller.

Improving the Inverse Dynamics Model of the KUKA LWR IV+ using Independent Joint Learning

Abstract: In this paper, we discuss the improvement of the inverse dynamics models of the KUKA LWR IV+ by a recently proposed approach called Independent Joint Learning (IJL). In IJL, the error between the torques from the real robot and the torques from inaccurate dynamics model is estimated using only joint-local information. Due to the reduced model complexity IJL can be used for task-to-task transfer learning and to a task different from the trained tasks. In this paper, we implemented IJL to improve the accuracy of the already existing KUKA LWR IV+ inverse dynamics model and our results show a significant improvement. We also discuss IJL for different types of input datasets and compared them in terms of performance.