Cost-Effective Calibration of Collaborative Robot Arm with Single Wire Encoder

Heungki Jeon, Martin B.G. Jun, Seung Han Yang, Huitaek Yun

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

With its flexibility of a collaboration robot and demands for maintaining virtual model’s fidelity in cyber-physical system of robotic manufacturing, frequent calibration is required. This paper proposes a cost-effective calibration method of a collaborative robot using a single wire encoder. First, DH (Denavit-Hartenberg) convention-based kinematics of the robot are analyzed. Then, a mathematical relation between distance errors between robot’s end effector and wire encoder and the kinematic parameters are presented. In the experiment, turning-in-place-based measurement methodology is applied. Also, in order to reduce error caused by wire bending at the point where a wire is hooked in previous studies, a rotatable wire hanger which has 2-DoF (Degree of Freedom) rotary joint is fabricated and applied. Next, identifiable parameters are analyzed through SVD (Singular value decomposition), then 21 parameters are calculated via Levenberg–Marquardt algorithm. The estimated parameters were validated by comparing fidelity of the calibrated and nominal model against actual physical system measured by a robot arm type measurement device. As a result, feasibility of the proposed method was checked by confirming 64.3% better fidelity of the calibrated model using the proposed method. Furthermore, the performance of the method also shows better results than camera-based robot’s internal calibration function provided by manufacturer.

Original languageEnglish
Pages (from-to)1615-1623
Number of pages9
JournalInternational Journal of Precision Engineering and Manufacturing
Volume24
Issue number9
DOIs
StatePublished - Sep 2023

Keywords

  • Closed-loop calibration
  • Levenberg–Marquardt algorithm
  • Robot calibration
  • Wire encoder

Fingerprint

Dive into the research topics of 'Cost-Effective Calibration of Collaborative Robot Arm with Single Wire Encoder'. Together they form a unique fingerprint.

Cite this