The motivation behind this thesis is to consolidate and evaluate the most common Proportional Integral Derivative (PID) controller tuning techniques used in industry. These are the tuning techniques used when the plant transfer function is not known. Many of these systems are poorly tuned because such consolidated information is not easily found in one single source such as this thesis. Once one of the tuning methods are applied almost always there will be further fine tuning needed to bring the system into the required design criteria. The purpose here is to find out which tuning technique will yield the lowest percent overshoot and the shortest settling time for all situations. This will give the engineer a good starting point; to minimally further adjust parameters to achieve the desired design criteria. There will also be discussion on the various algorithms used in industry. Four tuning methods will be evaluated based on their ability to control different style plants. The comparison criteria will be percent overshoot and settling time for an applied step input. The tuning methods chosen were the Ziegler-Nichols Open Loop method, the CHR method for 0% overshoot, the Ziegler-Nichols Closed Loop method, and the Rule of Thumb method. It is shown that for a second order plant with a lag and pure integration in its transfer function, the Open Loop method yielded the lowest results in terms of percent overshoot, yet the Closed Loop method had the shortest settling time. For systems of higher order than two it was shown that the CHR method gave the best performance however as the order increased the Closed Loop method gave a shorter settling time. For systems of higher order with varying lags in series the CHR method gave the best results. The Rule of thumb method usually gave similar results to that of the Closed Loop method; however for higher order systems the Rule of Thumb method gave less percent overshoot but with a longer settling time than the Closed Loop method. Since these tuning methods are used when the plant transfer function is not known, and none of the rules were found to give consistently the lowest percent overshoot, and settling time for all plants tested, there can not be a recommendation as to which method an engineer should choose to use. If the plant transfer function is known or can be reasonably modeled then the following recommendations can be followed. When tuning systems with pure integrations in their transfer function the Open Loop or Closed Loop method be used. When tuning systems of order higher than two the CHR or Closed Loop method should be used, however with high order systems with varying lags the CHR method should be used. It is the responsibility of the engineer to know how and when to implement each of the tuning rules properly.
Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd-4423 |
Date | 01 January 2007 |
Creators | Youney, Justin |
Publisher | STARS |
Source Sets | University of Central Florida |
Language | English |
Detected Language | English |
Type | text |
Format | application/pdf |
Source | Electronic Theses and Dissertations |
Page generated in 0.0017 seconds