Wrist camera-based vibration suppression control for a flexible manipulator

Xin Jiang, Atsushi Konno, Masaru Uchiyama

    Research output: Contribution to journalArticlepeer-review

    2 Citations (Scopus)

    Abstract

    This paper addresses a vision-based method for estimating vibration excited in the tip of a flexible-link manipulator. In this method, estimation of vibration is achieved by observing the variation of image features projected on a wrist camera. It mimics the situation of utilizing a wrist camera in tip vibration control of a space manipulator. In space, a vision sensor can be expected to be a feasible means for measuring the elastic vibration of the space manipulators, since they are more reliable compared with sensors like strain gauges. The method proposed in this paper takes advantage of the frequential characteristics of visual information that are reflected as a blurred background scene. With the high-frequency component of the projected image features, a Kalman filter-based observer is implemented as the estimator for the vibration. This implementation is characterized by the considerations of incorporating the slow sensor of the camera in the fast servo loop and compensation of the time delay due to image processing. With the vibration estimator, vibration suppression control relying solely on a wrist camera becomes possible. This scheme is successfully verified by experiments.

    Original languageEnglish
    Pages (from-to)805-823
    Number of pages19
    JournalAdvanced Robotics
    Volume25
    Issue number6-7
    DOIs
    Publication statusPublished - 2011 Apr 20

    Keywords

    • Flexible-link manipulator
    • Kalman filter
    • vibration estimation
    • visual servo

    ASJC Scopus subject areas

    • Software
    • Control and Systems Engineering
    • Human-Computer Interaction
    • Hardware and Architecture
    • Computer Science Applications

    Fingerprint Dive into the research topics of 'Wrist camera-based vibration suppression control for a flexible manipulator'. Together they form a unique fingerprint.

    Cite this