From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Neural network implementation

Hi to all, i'm completly new to neural network in labview and I have a problem.

I need to create  ANN whose input is signal that goes from 0 to 5V and the output is TRUE or FALSE depenging on whether the value of the signal goes above or under the treshold (1.5 for example...if signal value is above the output is TRUE, otherwise is FALSE). 
I know this is the simple problem but to me it seems so hard. 
In attachment you can see the part of the program that needs to be replaced with ANN.

Every kind of help is appreciated. 

0 Kudos
Message 1 of 10
(2,798 Views)

We cannot debug pictures.

0 Kudos
Message 2 of 10
(2,780 Views)

Oh i'm sorry, I totally forgot to attach .vi. Now I will do it. 

0 Kudos
Message 3 of 10
(2,713 Views)

Thanks for attaching the VI. Sorry, I have never used (or installed) the control&simulation module and have not experience with neural networks. Hopefully somebody can help you.

0 Kudos
Message 4 of 10
(2,695 Views)

Hi to all, I am a total newbie at fuzzy logic and I have a problem. 

I need to implement a fuzzy logic in my program instead of ""signal above treshold"" part. Fuzzy logic should have zero order hold value of my signal as an input and then it should generate either 1 or -1 value as an output depending on whether the zero order hold signal value is above or under the treshold(it could be 0.75 for example).

 

0 Kudos
Message 5 of 10
(2,703 Views)

Hi milosm,

 

why do you start a new thread for the very same VI?

Why do you need fuzzy logic now instead of neural networks?

 

I still don't know why you need both of them for your task of threshold comparison…

Best regards,
GerdW


using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019
0 Kudos
Message 6 of 10
(2,650 Views)

Hi GerdW, 

I need to use some kind of AI instead of above treshold comparasion and I figured that fuzzy logic wold be the most simple. I'm sorry for  posting about the same problem again but resolving this means a lot to me, it's for the msc degree.  I could't find at least some good tutorial for fuzzy in lv.  

Best regards ,

Milos

0 Kudos
Message 7 of 10
(2,639 Views)

Hi Milos,

 

to get a MSC degree you should be able to read some books on fuzzy logic, and to examine the example code coming with LabVIEW.

 


@milosm wrote:

I could't find at least some good tutorial for fuzzy in lv.  

The fuzzy control is part of "PID and Fuzzy control" toolkit. That toolkit also comes with a lot of example VIs on both subjects. And there are AppNotes on both subjects on NI's website…

You can also read Wikipedia or watch any videos on Youtube. (Or go the library of your university.)

Best regards,
GerdW


using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019
0 Kudos
Message 8 of 10
(2,631 Views)

@milosm wrote:

Hi GerdW, 

I need to use some kind of AI instead of above treshold comparasion and I figured that fuzzy logic wold be the most simple. I'm sorry for  posting about the same problem again but resolving this means a lot to me, it's for the msc degree.  I could't find at least some good tutorial for fuzzy in lv.  

Best regards ,

Milos


do you mean that one?

http://sine.ni.com/nips/cds/view/p/lang/de/nid/1394

 

it was once a quite expensive Toolkit, but is now included in the Full and Professional LabView IDE

 

it comes with some really well-written examples

http://zone.ni.com/reference/en-XX/help/371361R-01/lvpidmain/fuzzy_main/

 


@milosm wrote:

I need to use some kind of AI instead of above treshold comparasion and I figured that fuzzy logic wold be the most simple.


 fuzzy logic is a bunch of glorified thresholds 😉

 

 

in your first post, it looks like you are describing an activation function

(in the fuzzy toolkit, they call it membership functions)

and you want to threshold the output of the activation function.

 

if you want to start from scratch, you could exploit the hyperbolic tangent primitive .... 

 

 

 

Message 9 of 10
(2,627 Views)

Thank you alexderjuengere for the helpful articles. 

I came out with the solution after reading a few more tutorials, I guess I was just a bit lazy. 

For "mala vrednosti" i got the singleton output 0 and for the "veca vrednost"

 I got 1 which were just enough to control direction of the servo. 

 

0 Kudos
Message 10 of 10
(2,609 Views)