Mult-channel PID control using PID Toolkit

Question

I am controlling the light level in six locations (zones) using light sensors, dimmable ballasts (which respond to a 2 to 10 VDC signal) and the PID toolkit.  In my VI, I loop through 6 times (once for each zone) to control the light levels.  I use the simple PID.vi to determine the next setpoint for the dimmable ballast on the lights.  My problem is as follows, I think:
The PID algorithm uses past response and the current difference between the measured variable and the setpoint to calculate the next setpoint.  If my understanding is correct, then I am confusing the PID algorithm by calling it with 6 different data sets.  It is almost as if I need 6 separate PID functions, one for each zone. 
Perhaps I am wrong about the operation of the PID algorithm.  Can someone straighten me out?
Attached is a snapshot of the VI.  The FOR loop goes from 1 to 6, that is, once for each sensor zone.
Message Edited by rgentry on 04-29-2007 09:11 PM
Attachments:
temp.jpg ‏92 KB

Answer 1

rgentry,
The PID has two modes of operation: Single Channel and Muti-channel. The way you developed your algorithm, you are using single channel algorithm with multiple setpoint and variables. This will not work properly since LV will not keep the internal states correct for each channel.
To fix this, it is pretty simple. Remove the For Loop and provide the whole 1D array to the PID.vi. This will enable the "multi-channel" mode and every time you call your VI, it will keep the internal states for each channel, as if you had 6 different PIDs in parallel. Notice that the interpolation also will work the same way, you just need to remove the build array primitive.
Hope this helps!
Barp - Control and Simulation Group - LabVIEW R&D - National Instruments