Mpdel predictive control of a boiler

[PEintraining]

I may be misunderstanding things a bit, but it sounds as if you haven't overcome the steady-state error if your output drum level isn't settling at the correct value. Could you post some specifics about your model and your tuning approach? It looks as if your integral tuning could still use some work. All the details you can provide would be beneficial.

[syed334]

Thank you for ur kind reply,

I am using MPC to control DRUM LEVEL & DRUM PRESSURE of a BOILER model. I am not using any TOOL BOX OR COMMAND FUNCTION like MPCobj, i have made my own COST FUNCTION n other necessary m flies. I am considering my input STEAM FLOW rate as CONSTANT DISTURBANCE, WITH DISTURBANCE my MPC was unable to track OUTPuT to the reference value ZERO for LEVEL & PRESSURE 7 was having steady state error. So, I have AUGMENTED MY A,B,C matrices by INTRODUCING INTEGRAL ACTION. AFTER introducing INTEGRAL, DRUM PRESSURE IS REGULATING TO ZERO, BUT DRUM LEVEL REMAINS the same as before without INTEGRAL ACTION. Thus, any technique to overcome it, I have applied MPC+FEEDFOWARD to reject disturbance, but still DRUM LEVEL is having STEADY STATE ERROR, fOR YOUR REFERENCE I am using ASTROM & BELL DRUM BOILER MODEL. I hope it is clear now, hope to have a suggestion. Thus could have give me any idea for coding FEEDFORWARD+MPC to reject disturbance so that Drum level regulates to the desired level or any paper in which it is mentioned clearly.