Prev: unable to debug
Next: Problem with updating a Simulink modell from ML 13 to 14
From: Nazar on 18 May 2010 09:42
I seem to have run into a conundrum of sorts using the GA; hopefully some of you can help me out.
There are examples for custom populations being either integers or specific user defined random values within geometrical shapes and such; however, I seem to run out of coding ideas when it comes to restricting a possible population (for one given variable) to a localized area.

Shortly said, I need an initial population to spawn within a predefined area and never leave that area.

In this situation the best way to describe the area is:
Square S1 with side length a1 = 1
Square S2 with side length a2 = 0.6

The limiting area being the subtraction of S2 within the S1 area, both being centered at the same point.
(Think of a picture frame)

Now, this being said; here is where I run into a wall:

1 - Is it possible to program that specific shape while respecting the proper syntax for inequality conditions of the GA function?
2 - If not, what alternatives do I have?

Now, I am not expecting a full answer ( not that I would mind :) ), but I would appreciate a couple of ideas.

Thank you
Nazar


PS: I did not find any similar questions out there; if I missed one, please link to it :)
From: Nazar on 18 May 2010 09:48
I apologise, slight correction:
This is for two variables ofcourse (not one given variable as specified previously)
meaning both on x and y.


"Nazar " <nazar.delegan(a)gmail.com> wrote in message <hsu5fc$q6h$1(a)fred.mathworks.com>...
> I seem to have run into a conundrum of sorts using the GA; hopefully some of you can help me out.
> There are examples for custom populations being either integers or specific user defined random values within geometrical shapes and such; however, I seem to run out of coding ideas when it comes to restricting a possible population (for one given variable) to a localized area.
>
> Shortly said, I need an initial population to spawn within a predefined area and never leave that area.
>
> In this situation the best way to describe the area is:
> Square S1 with side length a1 = 1
> Square S2 with side length a2 = 0.6
>
> The limiting area being the subtraction of S2 within the S1 area, both being centered at the same point.
> (Think of a picture frame)
>
> Now, this being said; here is where I run into a wall:
>
> 1 - Is it possible to program that specific shape while respecting the proper syntax for inequality conditions of the GA function?
> 2 - If not, what alternatives do I have?
>
> Now, I am not expecting a full answer ( not that I would mind :) ), but I would appreciate a couple of ideas.
>
> Thank you
> Nazar
>
>
> PS: I did not find any similar questions out there; if I missed one, please link to it :)
 | 
Pages: 1
Prev: unable to debug
Next: Problem with updating a Simulink modell from ML 13 to 14