How can I choose optimal signal?

[nguyenmanhcuong]

Hi all!
I received some signals from device. Also I've had a known standard signal. My problem is: how can I choose one (of the receiceved signals), which compares with standard signal with the smallest error?

In the picture attached here, yellow graph in bold  is reference (standard). How to make a decision: which one of remaining graphs is closest to the reference graph? 

Thank for your any helps!

Mc

[AnalogKid2DigitalMan]

Mc:

Your graphs did not attach.

-AK2DM

[nguyenmanhcuong]

oh, I'm sorry. Here is attached graph.

[Meghan R]

Hello Mc,

Can you give more detail on what you're attempting to do?  Are you wanting to programmatically choose the plot that is closest to the reference?  If that is the case, you have quite a few options on Functions Palette, including an Express VI that does polynomial fitting.  Please let me know if you need further assistance.

Thanks!

Meghan 

Applications Engineer

National Instruments

[nguyenmanhcuong]

Hello Megha,

Yes, I'm wanting to programmatically choose the plot that is closest to the reference? Out put of the task is this plot. But I haven't found polynomial VI, which can solve this problem.

Thanks!

Mc

[Henrik_Volkers]

First you have to define you best match. (closest fit, same force at a specific point, same decreasing rate of force .... ??)

Your signals look quite simple, so a MSE calculation might workout best. 

Even more simple is the sum of abs difference 

 

Message Edited by Henrik Volkers on 04-02-2009 09:05 AM

[nguyenmanhcuong]

Hello Henrik!

Thank for your advice, but your program could not run on my Labview 7.1. How should I do now?

[Henrik_Volkers · Accepted solution]

That's why I post the picture , I don't have 7.1 (any more)  ,  this answer took longer to write than to code this thing new 😉

the mse.vi is located in the mathematics- statistic section 

the rest is  at numeric, array and waveform

PS: the help section and the search functions  are the tools to look at 

       Kudos welcome 

[nguyenmanhcuong]

Many thanks for all. I have solved problem.

Mc