PLC5 PID Loop help

[sparkotronic 0]

Hi, I've been looking at a part of one of the processes in my work place and have came across something I have never seen before. Here is a simplified drawing of the process The PID loop controls the burner output to maintain the temperature at T1. Material is fed into the system at a fixed speed by the feed belt. The material is then sucked through the drying chamber and deposited in a silo by induced draught. On the PID loop there is a variable called the Bias. This is changed by a calculation (%feedbelt speed - 50 X 0.4) So if the operator sets the speed of the belt to anything other than 50%, the bias value changes. The belt normally runs at 60(ish)%. This is a fixed speed set by the operator and isn't normally adjusted unless on a product change. My questions are what is the Bias and what effect does varying the Bias have on the loops performance. I have a very limited experience on PID loops, I did phone Rockwell but they couldn't keep it sparky simple :) Thanks in advance

[Mickey 71]

Bias or feedforward is used to bump the output when changes to the process occurs. In your case the speed of the conveyor. When the speed slows down the bias will reduce the PID's output, (not as much heat is required) before the PID sees a change in temperature T1. The opposite is true when the conveyor speeds up. This increases the response time of the PID control. See links below for more on "Feedforward" (Bias) http://blog.opticont...om/archives/297 http://www.google.co...iw=1584&bih=807 Edited 8 Mar 2013 by Mickey

[sparkotronic 0]

Hi, Say for instance if the conveyor is running at 62% the bias has a constant of 4.8. Does this increase in output stop when the loop catches up?

[Mickey 71]

No the bias is still there but the PID instruction is in control of the output. In others words the PID is going to do whatever it has to do to maintain setpoint. The bias or feedforward is used to kick start the output in anticipation of a change in the process variable ( temperature in your case.) Edited 9 Mar 2013 by Mickey

[sparkotronic 0]

Thanks for clearing that up for me, it makes much more sense now Cheers, Colin

[Mickey 71]

Your welcome. Edited 9 Mar 2013 by Mickey