PID loops

A

Anbu Selvarajan

Yes... I too agree with David.
I shall add few more ideas on to this..

Generally Derivative action in PID controller takes rate of change of error, i.e. d(e(t))/dt
The sudden change in the output of the PID controller for a step change in the setpoint due to derivative action, what we usually call "Derivative kick".

This "Derivative kick" can be avoided by computing rate of change of pv, i.e. d(pv(t))/dt instead of conventional method of finding rate of change of error in PID controller.

'Cos the PV, the process output won't respond suddenly to any change in the setpoint due to its inherent lag (time constant).
 
> I think you can use PID controller with autotuning function. In this method your controller changes and plays with PID parameter and achieves the best parameter depending on your system specification. <

---- snip ---

I don't agree with the suggestion of auto-tuning controllers. In my experience (over 20 years) I've never had good results with auto tuning. Usually the settings obtained by auto tuning are much too aggresive. If there is not too much time lag in the system between control valve movement and system response then the ZN tuning is the way to go.
 
Top