Dear Parag Patil ,

Many modern thermal processing systems involve temperature control of heated plates. Typically, these plates are heated by infrared radiation from hot filaments (including tungsten halogen lamps) and the temperature is measured using pyrometers. In many cases the temperature of the system is controlled using a Proportional-Integral-Derivative (PID) controller. However, there are applications where the temperature must be ramped up (or ramped down) rapidly while maintaining good temperature uniformity and with tight performance limits. In these cases, it may be difficult, or even impossible, to achieve the desired performance using PID controllers. In addition, the dynamic response of these systems typically changes considerably with operating conditions such as temperature, process gas composition, wafer emissivity, or wall emissivity (e.g., in systems where walls become coated during the deposition process). These changing operating conditions can make it difficult to tune a PID controller to achieve the desired performance over a broad range of operating conditions.

https://www.scsolutions.com/controls/projects/model-based-versus-pid-control/

Regards,

Shafagat

Hi Parag Patil ,

The PID controller is based on an 'estimated transfer function' of the process. So the actual performance of the process with the chosen PID controller would depend on the following:

You might want to look up 'Relay Auto-tuning' methods for PID controllers.

Best,

Sujay

Sujay Dilip Kadam Thank you for your suggestions

Thank you Professor Michael Short for your suggestion.

I would like to reiterate your suggestion as per my understanding (Please correct me If I am wrong ):

The parameters of PID-controller based on detrended / not-detrended I/O data should be the same (provided, the transfer function estimated should be sufficiently capturing the process ).

I do not need to adjust the controller parameters (although the data used to tune the controller is detrended).

Thanks a lot for your suggestion.

Short answer; yes you can use your detrended model for PID tuning. You just have to be careful of adding appropriate integral action in the PID.

Longer answer; what ever you don't include in your model when desiging a controller, you have to consider it an external disturbance. In your case, by ignoring the trend in your model, then you have to design you controller to compensate for slow disturbance and offset. Integral action is what is used to compensate for such kind of disturbances.

Othe examples of unmodelled disturbances are gain and delay, For example an unmodelled gain disturbance can be a process that has different gains on different operating points. You decide to ignore the different gains and use an average of all gains in your model for control. Then you design over this average model, but you have to be careful to have good gain margins in the design, as to be robust against the gain variations that your real process has.

I recommend you to use this open source tool for PID tuning:

https://pidtuner.com