Complex function optimization

[navneet]

Hello everyone,

There are very few examples on optimization which exists for using the Nonlinear Optimization subVI. Anyone has tried to perform optimization of multi variable complex function using the available subVI? Care to share some insights and examples?

cheers,

[Jesse_D]

Hi navmeet, 

Am I right in assuming that you are referring to the "Constrained Nonlinear Optimization.vi" in the Optimization palette? If so there is an included example in the detailed. 

I really like math problems, if you give me an idea of what you are trying to do perhaps I can help you on your way. 

[navneet]

Thanks Jesse,

I have a complex function and trying to find parameters estimates. The problem is beacuse of its non-uniqueness it can give many solutions, and there is always a chance of being trapped by a local minima. Perhaps I can email you a document identfying the problem?

Then rest we can continue here, if you want.

Thank you very much for offering to help Jesse 🙂 

[Jesse_D]

Hey navmeet, 

That sounds great, I will send you a PM. 

[Jesse_D]

Hi navmeet, 

I didn't hear from you. Did you get my pm? 

[navneet]

Hi Jesse,

Sent you a pm with the formula. Its in basic shape, but I would be formulating for extra terms. I just want to know which will be a good starting point to build what i am trying to do :). Once I understand the starting point (subVI's, structures needed) I will have no problem. Currently am studying some algorithms developed by various workers and trying to see how it can be implemented with LabVIEW. 

[Jesse_D]

Hi navmeet, 

Here is the code I was playing with to determine that the optimization function only finds local minimums. Depending on where you start the optimization you will find different local minimums in the test function I made.

[navneet]

Hi Jesse,

Thanks for the example. Sorry for getting back late, got caught up in other things simultaneously.

I will have a look the the example and see what can be built on 🙂

Will get back to you soon.