This is the code for the simulations of multi-agent systems done for a research group. The simulations are of an ant-colony that is flexible enough to adapt its resources of ants on the fly when foraging for food. Ants are nearly blind, and rely on pheromone trails left by ants before them to find the best trails to food. These pheromones accumulate over time, and the problem is to ensure that the trails that lead to the best and closest food sources accumulate the most pheromones, and to ensure that this is true after closer food sources are suddenly made available.
I was responsible for the design and implementation of the algorithms behind the ants and logic for the simulation. The Renderer was written by a colleague.
Our work was published in the 2017 Modern Artificial Intelligience and Cognitive Science Conference.
S.Kazadi, G. Jeno, X. Guan, N. Nusgart, A. Sheptunov. Decision making swarms. Proceedings of the 28th Modern Artificial Intelligence and Cognitive Science Conference, Fort Wayne, Indiana, 2017.
Here is the slide deck that I used to present our work at the conference.