#4 Add more ArtificialAnt-based problems

Medium_Priority
open
nobody
5
2005-06-19
2005-06-19
Emperor XLII
No

The structure of the ArtificialAnt problem allows for many
variations. Currently, all examples deal with the exploration-
style problem (path following, finding food). Other
possibilities and general enhancements include:

- Food gathering (Find food and carry it back to a goal, a.k.a.
“ant hole”)
- Obstacle avoidance (Finding a path to a goal while avoiding
obstacles, such as navigating a rock-strewn field, or
navigating a maze)
- Food organization (Find food and move it into a pattern,
such as a single clump (no predifined spot), or dispersed
equidistantly)

- Cooperative behavior (This can be used with several of the
problems described above. It involves instantiating multiple
Individuals and using their cumulative score in the
Environment to calculate the fitness of that Individual class.)
- Trail making (Creating “pheromone trails” for use in the
above problems, which can make return trips to a food spot
or a gathering point easier for the ants)

These may be implemented as separate Individual classes,
though the same environment could probably be used with
minor changes that would not interfere with the current
version.

Discussion