• DocumentCode
    807065
  • Title

    Framework for reducing digital-to-analogue converter quantisation error in servo control systems

  • Author

    Yeh, T.-J. ; Lu, N.-S.

  • Author_Institution
    Dept. of Power Mech. Eng., Nat. Tsing Hua Univ., Hsinchu
  • Volume
    2
  • Issue
    8
  • fYear
    2008
  • Firstpage
    693
  • Lastpage
    705
  • Abstract
    A framework for reducing the digital-to-analogue convertor (DAC) quantisation effect in servo control systems is proposed. In the framework, two compensators are used to shape the spectrum of the DAC quantisation noise. One compensator, which is an anti-windup-based compensator, not only filters out the high-frequency part of the quantisation noise, but also provides stability and performance in the presence of control saturation. The other compensator, which is a quantisation error feedback compensator, provides effective reduction of the low-frequency noise and can account for the spectral distribution of DAC quantisation noise by an optimal choice of some parameters. Theorems concerning the solvability and the stability are given. Systematic procedures to design the compensators for single-input-single-output (SISO) and multi-input-multi-output (MIMO) plants are also suggested. To validate the performance of the framework, servo control of an optical disc head is considered. Experiments indicate that the output error caused by DAC quantisation is significantly reduced.
  • Keywords
    digital-analogue conversion; optical storage; servomechanisms; DAC quantisation noise; anti-windup-based compensator; control saturation; digital-to-analogue converter quantisation error; low-frequency noise; multi-input-multi-output plants; optical disc head; quantisation error feedback compensator; servo control systems; single-input-single-output plants; spectral distribution;
  • fLanguage
    English
  • Journal_Title
    Control Theory & Applications, IET
  • Publisher
    iet
  • ISSN
    1751-8644
  • Type

    jour

  • DOI
    10.1049/iet-cta:20070165
  • Filename
    4567171