hi I spent my time reading books and articles about the ants behaviour. My goal is to simulate an ant colony behaviour. I still have some questions and I hope that someone could help me. I would like to create a model(Finite state machine) about how the ants collaborate together to retrieve a prey. basically, the steps of the ants behaviour are the following: an ant try to move a prey. if unsuccessful, it releases a pheromon for recruit nestmates which are close to the prey. try to move the prey. if unsuccessful, it goes to the nest depositing a pheromon from the prey to the nest. I have some difficulties to know: how they are able to find the nest? if they can move the prey, they have to go the nest. But if there is an obstacle they have to change of direction. how are they able to decide to go away from the nest? thank you for any help you can give!!!!

IIRC, Ant's lay pheromon trails from their nest to wherever they go.

Ant's also somehow know to use the "least difficult path" rather than the shortest path (straight line)

[edited by - caffeine on June 19, 2002 3:39:25 PM]

About a month ago there was a good topic on this. Check my profile and look if it is in my recently replied to list. Otherwise, use the search thing.

Trying is the first step towards failure.

hi,


I didn''t find what I am looking for on this forum. But it was a nice idea ragonastick. So I am always looking for information about the way to modelize the ant behaviour in the transport task. I found several experiments about robots but nothing about ants itself.

Nobody has suggestion?????????

If I were attempting this project, I would try use some form of behavioural modification for each ant based on adaptation.

This may not yield the best results but would definitely be interesting.

This would change the dynamics of the problem.
Instead of trying to determine a working FSM yourself, you would be attempting to find some form of encoding of the FSM for each ant so that adaptation could take place (ie. a ''genome'' for the ant).
This might be an advantage because you let the fitness of each ant determine the ''goodness'' of the FSM.
One thing to note is that the overall fitness would probably need to be determined at a ant-colony level (not ant level) if you want some form of communication between ants for emerge.

Just a thought!

errrr... didnt we have exactly the same thread just a little while ago ??? I dont have the time to retype ALL the stuff I said, so could you refer to the previous thread for my thoughts

---

Sancte Isidore ora pro nobis !