LQP

[KF_]

Hi,

I have a question about the LQP block in Labview. Does it always give a solution even when it comes across something non-feasible? 

Kind regards

KF_

[muks]

I am sorry but can you expand LQP?

[KF_]

Hi

Sorry for that: its linear quadratic programming, its located in the mathematics section under optimization with name QP. And there are 2 ways to use this block with interior point method or active set method. I am interested in active set method.

Kind regards

KF_

[ThiCop]

It depends on what you consider to be non-feasible:

- For example if you use A=0 and B=1 for the constraint A * x=b then your QP(AS) VI produces an error.

If you want to see for other infeasible constraints, then you can always try it yourself or is this not possible?

If my interpretation of infeasible was incorrect don't hesitate to let me know.

[KF_]

Hi Thicop,

I want to use the QP block for Model predictive control, so it is continously solving problems, but it can come across an infeasible problem once in a while if you use hard constraints. An advantage of the active set method is that when you come across an infeasible problem, it can just return the previous solution, because active set method always starts from the previous solution set. But I guess this block doesnt do that.

Best regards,

KF_

[ThiCop]

Hello,

There's a warm start option for QP: AS  ( see http://forums.ni.com/ni/board/message?board.id=170&message.id=487059 ) , so that " LabVIEW uses the indexes of the active constraints from the previous solution as the initial set of active constraints for the current problem". You just ahve to enable it.

Have you tried what happens with an infeasable problem yourself? Or  is that not possible for you?

[KF_]

Hi,

I guess the warm start is what I need, I'll try it and see what i get!

Regards

KF

[ThiCop]

Hello,

When you have any result (bad or good) please let us know.

Good luck with the programming.