Enhancing Classical Impedance Control Concepts while Ensuring Transferability to Flexible Joint Robots

Abstract

In robotics the concept of impedance control enables a controlled and therefore safe interaction behavior between robot and human. The objective of this bachelor’s thesis was to evaluate several configurations of the enhanced Elastic Structure Preserving Impedance (ESPi) controller. The enhanced controllers reduce the maximal required time-derivative order of the link position by one, compared to classical approaches. Four different configurations, representing different mass-spring-damper designs,were transformed in a three degrees-of-freedom form, simulated and numerically optimized for improved noise damping and disturbance rejection behavior. The H-infinity method was used. The configuration ESPi2 showed the best noise damping in simulation. The results on the testbed confirm a robust implementation, improved noise damping and reduced control effort for disturbance rejection over the initial controller.