Accurate calibration of systematic errors for car-like mobile robots using experimental orientation errors

Daun Jung, Jihoon Seong, Chang bae Moon, Jiyong Jin, Woo Jin Chung

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

One of the essential technologies for autonomous navigation is localization. Localization is important because accurate pose estimation is required for path planning and motion control. In order to improve localization accuracy, a relative positioning method on the basis of accurate odometry is necessary. Odometry calibration methods for two wheel differential drive robots have been researched for many years. However, it is difficult to find odometry calibration methods for car-like mobile robots. In this paper, an accurate calibration method for car-like mobile robots is proposed. Experimentally measured orientation errors were used to improve the accuracy of the calibration method. There are two contributions in this paper. The first is the significant reduction of calibration errors by the use of accurate calibration equation. In the previous research, calibration equation required approximations. However, there is no approximation error in the proposed equation owing to the use of orientation errors, not positional errors. The second is the experimental convenience. The orientations can be easily measured by onboard inertial sensors. Therefore, calibration experiments can be easily carried out in both indoor and outdoor environments. The presented experimental results show that resultant performance of the proposed scheme is superior to the results of the previous research.

Original languageEnglish
Pages (from-to)1113-1119
Number of pages7
JournalInternational Journal of Precision Engineering and Manufacturing
Volume17
Issue number9
DOIs
Publication statusPublished - 2016 Sep 1

Keywords

  • Calibration
  • Localization
  • Mobile robot
  • Odometry

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Industrial and Manufacturing Engineering
  • Mechanical Engineering

Fingerprint Dive into the research topics of 'Accurate calibration of systematic errors for car-like mobile robots using experimental orientation errors'. Together they form a unique fingerprint.

  • Cite this