Description

kevindekemele

opened on Apr 22, 2021

Issue body actions

I'm plotting the Nyquist diagram of a transfer function which has poles on the imaginary axis:

import control
import matplotlib.pyplot as plt
import numpy as np

K = 1

Gp =  control.tf([1], [1,1,0,0,0]) 
Gc = control.tf([K], [1]) 


control.nyquist_plot(Gp*Gc,plot=True,omega_num=1000)
plt.show()

Which results in the following incorrect plot, that does not show infinte semicircles, but something incorrect:

Also using indent_direction and indent_radius did not solve this.
The trick I found that makes it work is if I add omega_limits=[0.001,10000] to nyquist_plot:

Can you please fix such that the incorrect 'line' does not appear?

Activity

1. bnavigator commented on Apr 24, 2021

   

   bnavigator

   on Apr 24, 2021

   Contributor

   More actions

   Yes, the algorithm to determine the default omega vector for the system starts with a lower frequency of 0.1, because the lowest non-zero pole of the system is at -1.

   python-control/control/freqplot.py

   Lines 1088 to 1092 in e81f648

   	# This code looks at the poles and zeros of all of the systems that
   	# we are plotting and sets the frequency range to be one decade above
   	# and below the min and max feature frequencies, rounded to the nearest
   	# integer. It excludes poles and zeros at the origin. If no features
   	# are found, it turns logspace(-1, 1)

   The "trick" to manually define the omega limits is the correct way to override the default omega vector determination.
2. kevindekemele commented on Apr 25, 2021

   

   kevindekemele

   on Apr 25, 2021

   Author

   More actions

   Perhaps if there is a pole at the origin, the lower limit of the omega vector should be below the indent_radius ?

   Reacted by Richard Murray
3. murrayrm

   linked a pull request that will close this issueRefine automatic contour determination in Nyquist plot #620on Jun 5, 2021
4. murrayrm

   closed this as completedin #620on Jun 16, 2021

Issue actions