Difference between revisions of "Ziegler-Nichols tuning rules don't quite match the original gains from the Ziegler-Nichols paper"

From FBSwiki
Jump to: navigation, search
Line 15: Line 15:
  
 
* For the frequency response method, the proportional gain for a PI controller should be k_p = 0.45 k_c (instead of 0.4 k_c and the integral time constant should be T_i = T_c/1.2 (instead of 0.8 T_c).
 
* For the frequency response method, the proportional gain for a PI controller should be k_p = 0.45 k_c (instead of 0.4 k_c and the integral time constant should be T_i = T_c/1.2 (instead of 0.8 T_c).
 +
 +
These changes also impact the gains listed in the subsection on the [[Original Ziegler-Nichols gains listed after improved tuning rules are incorrect||improved Ziegler-Nichols tuning rules]].

Revision as of 08:45, 3 September 2019

Edition Second
Version number 3.0j
Submission date 31 Aug 2019
Your name Richard Murray
E-mail address murray@cds.caltech.edu
Chapter 11
Page number 11-12
Line number 8
Current status Closed

The gains listed for the original Ziegler-Nichols tuning rules don't quite match what was in the original paper:

  • For the step response method, the integral time constant for a PI controller should be T_i = \tau/0.3 (instead of 3 \tau).
  • For the frequency response method, the proportional gain for a PI controller should be k_p = 0.45 k_c (instead of 0.4 k_c and the integral time constant should be T_i = T_c/1.2 (instead of 0.8 T_c).

These changes also impact the gains listed in the subsection on the |improved Ziegler-Nichols tuning rules.