We discovered an other project that aims to simulate ants behaviors, which is called Myrmedrome. It is based on some behaviors like pheromones of courses, but also trophalaxis, it differentiates soldiers and workers, and also allows to put insect preys on the map, and to wreck havoc within ants with a finger.
But it is not related to scientific papers like our project will, ours also aims for computer science applications (at least from far away). And in our simulator, ants move freely around the map (and outside), while in Myrmedrome, they follow a tile-based map.
It also appears that the ants in Myrmedrome are not likely to improve path to food by themselves, and will rather follow the first defined path, even not optimal.
So it gives us an aim : to make our project better than this one !