제어로봇시스템학회:학술대회논문집
- 2002.10a
- /
- Pages.62.3-62
- /
- 2002
New model reduction method and optimized the Smith predictor disign using reduced model
- Jeoung nae choi (Wonkwang Univ.) ;
- joon ho Cho (Wonkwang Univ.) ;
- Hwang, Hyung-Soo (Wonkwang Univ.) ;
- Park, Moon-Soo (Jeonju Tech. College)
- Published : 2002.10.01
Abstract
In this paper, we proposed a control technique that can be applied to various processes. The most of the process can bereduced to second order plus time delay (SOPTD) model. And we proposed improved model reduction algorithm using geneticalgorithm. This method considered four points to reduce the error between original model and reduced model in the Nyquistcurve. And, to compensate time delay, the Smith predictor plus PID controller is adopted. And a new PID tuning algorithm wasproposed, which got from numerical analysis and can be obtained the optimal performance. The PID parameters are obtainedfrom the coefficients and time delay of reduced model. The simulation results show the validity.
Keywords