본문 바로가기
  • Home

A Study on Controller Design for Time Delayed System

  • Journal of Knowledge Information Technology and Systems
  • Abbr : JKITS
  • 2017, 12(6), pp.879-885
  • DOI : 10.34163/jkits.2017.12.6.009
  • Publisher : Korea Knowledge Information Technology Society
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Published : December 31, 2017

Joung, Houng Kun 1 Lim, Dong Kyun 1

1한양사이버대학교

Accredited

ABSTRACT

This study compared methods of designing a robust PID controller, which satisfies the design specification in a system with time delay, and has robustness and stability at the same time. This study compared the PID controller tuning method of Ziegler and Nichols and the PID controller design method by using the suggested by Grassi of . The PID controller tuning methods of Ziegler and Nichols include a step response method, a frequency response method, and a modified method. The PID controller tuning methods of Ziegler and Nichols are widely used in industrial fields and are easy to design mathematically. They are PID control tuning methods widely used for the systems that the dynamic characteristics of a plan are not fully known. The PID controller design method using Grassi of is a PID tuning method that norm, which guarantees the stability and robustness, uses the convex optimization method. This study compared the method of Ziegler and Nichols and the method proposed by Grassi of to propose a controller tuning method guaranteeing the stability and robustness, supposed to be guaranteed by a PID controller. Moreover, the performance of two methods was compared through simulations. In order to consider the design specifications for the frequency domain, simulations were compared after dividing the design conditions into a low-frequency barrier, satisfying the command follow-up and external disturbance canceling performances, and a high-frequency barrier, corresponding to the low sensitivity performance against the sensor noise.

Citation status

* References for papers published after 2023 are currently being built.