Ziegler Nichols Tuning Method - Danfoss VLT AutomationDrive FC 300 Design Manual

3. Introduction to FC 300
3.
4.
3

3.3.5. Ziegler Nichols Tuning Method

In order to tune the PID controls of the frequency converter, several tuning methods can be used.
One approach is to use a technique which was developed in the 1950s but which has stood the
test of time and is still used today. This method is known as the Ziegler Nichols tuning method.
The criteria for adjusting the parameters are
based on evaluating the system at the limit of
stability rather than on taking a step response.
We increase the proportional gain until we ob-
serve continuous oscillations (as measured on
the feedback), that is, until the system be-
comes marginally stable. The corresponding
gain
od of the oscillation
period) is determined as shown in Figure 1.
P
should be measured when the amplitude of oscillation is quite small. Then we "back off" from
u
this gain again, as shown in Table 1.
K
is the gain at which the oscillation is obtained.
u
Type of Control
PI-control
PID tight control
PID some overshoot
Table 1: Ziegler Nichols tuning for regulator, based on a stability boundary.
Experience has shown that the control setting according to Ziegler Nichols rule provides a good
closed loop response for many systems. The process operator can do the final tuning of the control
iteratively to yield satisfactory control.
38
Set par. 7-34 (Integration Time) to 20 sec. and reduce the value until the feedback signal
again begins to vary continuously. Increase the integration time until the feedback signal
stabilises, followed by an increase of 15-50%.
Only use par. 7-35 for very fast-acting systems only (differentiation time). The typical
value is four times the set integration time. The differentiator should only be used when
the setting of the proportional gain and the integration time has been fully optimised.
Make sure that oscillations on the feedback signal is sufficiently dampened by the lowpass
filter on the feedback signal.
NB!
If necessary, start/stop can be activated a number of times in order to provoke a
variation of the feedback signal.
NB!
The method described must not be used on applications that could be damaged by
the oscillations created by marginally stable control settings.
(K ) is called the ultimate gain. The peri-
u
(P )
(called the ultimate
u
Proportional Gain
K
0.45 *
K
0.6 *
u
0.33 * K
MG.33.B9.02 - VLT
Illustration 3.2:
system
Integral Time
P
0.833 *
u u
P
0.5 *
u
0.5 * P
u u
® is a registered Danfoss trademark
FC 300 Design Guide
Figure 1: Marginally stable
Differentiation
Time
-
P
0.125 *
u
0.33 * P
u