Home /Research /Disturbance Observer Based Control for Quasi Continuum Manipulators
MANIPULATION

Disturbance Observer Based Control for Quasi Continuum Manipulators

Daniel Müller, Carina Veil, Oliver Sawodny

Year
2020
Citations
13

Abstract

Nowadays, robots are an essential part of modern production lines, usually working in a designated area since they can pose a threat to human workers. The so-called soft robots constitute a human-friendly alternative to classic industrial robots, even allowing for human-machine collaboration. This is possible due to their soft and therefore inherent safe structure. In this paper we consider quasi continuum manipulators (QCMs), a special kind of soft robots. Their dynamic behavior is affected by friction as well as their soft materials. Dynamical models are thus hard to identify, suffering from imperfections and uncertainties. To overcome these flaws we propose a disturbance observer (DOB) based controller using an extended Kalman filter (EKF). We show superior performance on a real robot compared to an existing benchmark concept based on a PID-like controller. The generalization of this approach is demonstrated by implementing our method on two QCMs with different kinematics.

Keywords

RobotControl theory (sociology)Control engineeringKinematicsExtended Kalman filterComputer scienceGeneralizationPID controllerBenchmark (surveying)Kalman filter

Related papers

Browse all MANIPULATION papers