PID Lead Lag Compensator Units

[falexander]

Hello,

I have Real Time PXI Target and few PID loops running using the LabVIEW PID and Fuzzy Logic Toolkit. However, the system that i am trying to control is unstable. The system can be made stable by using lead lag compensator. I have already tried the lead lag compensator using MTS Controller and the system works great. But, I have to switch back to LabVIEW PXI Target. So if i can implement the same lead lag filter in LabVIEW, the system can be made to work. 

The problem is that in LabVIEW, the PID Lead lag asks for the units in miuntes (time). 

http://zone.ni.com/reference/en-XX/help/370401D-01/lvpid/pidlead_lag/

http://digital.ni.com/public.nsf/websearch/4265D666730AFC4D86256CD300812E80?OpenDocument

I have not been able to find out how to interpret what this means or how to convert this units to Hz or radians. Basically, I need to convert the Lag time (minutes), Lead time (minutes) to poles (Hz) and zeros (Hz).  After that, I can implement this in the PID loop and make the system work. So please help...

Thank you, 

Alexander

[Justin__D]

Hi Alexander,

So you are trying to implement this lag-lead vi within your PID loop, correct?

[falexander]

Yes, I have to implement this Lag-Lead VI within the PID loop. I need two of this  Lag-Lead VI, one for the lag compensator and one for the lead compensator. 

The PID parameters (P, I, and D) units are also in minutes, but i didn't care about the units for this. I just tuned it to follow my command. 

But, with  Lag-Lead compensator,  I know the parameters (the location of the poles and zeros) in frequency. But, how do i convert those units to minutes (time)? Or how may I interpret the LabVIEW units?

Thank you,  

[Justin__D]

Hi Alexander,

I have found a link that I think explains this pretty well.  Does this help?

Control Systems/Controllers and Compensators

[falexander]

Well, I have some text books that explains all these compensator, etc... But, I have not been able to find out how to convert the location of poles and zeros to lag time and lead time in minutes (to enter into LabVIEW) in any text book or any online resources (including that link). BTW, I don't know about that site because some of what they state doesn't seem to be correct. 

I really needed to use the lead lag compensator. So I thought I would try it to figure it out, and I did. I will write what I did so If it ever helps anybody in the future....

I made a loop with analog inputs and analog outputs in LabVIEW. Then I used the PID Lead lag VI and the LabVIEW Graphs. Then I used a spectrum generator to input random noise (which contains all frequencies in the range that you specify) into the Lead Lag VI. I used the output of the Lead Lag VI to send the outputs to the spectrum generator. I set the spectrum generator to do  a closed loop analysis and give results in bode plots. Then I would change the lag time and lead time parameters and see how the bode plot changes. From this, it can easily find out the poles and zeros. Since I know the location of the poles and zeros, I would change the lag time and lead time parameters until the bode plot in the spectrum analyzer gives the bode plot that I am looking for.  I also used a frequency generator and a scope to see the signal attenuation and phase shifts at various frequencies, and it is as it should be.  Well, hope this helps..

Alexander