[Swarm-Modelling] new ant system design/implementation - first notes

[Nelson Faria]

Hi all,

I started a implementation of an ant system.
The main goal for this system is to put the ants
solving complex labyrinths in research of food.

The map is a grid like a chess board, with
inaccessible parts (labyrinth like).

My first implementation I introduce some parameters
like:
- every ant left different pheromones on terrain
- the ants ignore the own pheromone, but "smell" all
the pheromones from the others.
- by time the pheromones are clean from terrain.
- the ants left different quantities of pheromones
when  carry food.
- the ants have a extra parameter to decide where to
go, that is the current direction.

the current work don't have any obstacles, and no
food, only to see the ant colony behaviour. 

The main problem I see on my implementation is:
- some ants(2-3 ants) start walking in circle and
follow the other pheromone ants.

How can I avoid this problem?
If you have any ideas, please share :)

Nelson Faria



                
__________________________________ 
Do you Yahoo!? 
Yahoo! Small Business - Try our new resources site!
http://smallbusiness.yahoo.com/resources/