I worked on a clustering model that follows a decentralised reaction-diffusion-chemotaxis aggregation scheme. This aggregation scheme is inspired from a biological phenomenon seen in slime mold. I used a multi-agent approach simulated on a 2D cellular automaton. The automaton is used as an environment for agents to communicate and move. The challenge was to make a rudimentary model where the clustering task should emerge from simple rules governing the evolution of automaton’s cells.
Skills learned :
- Mathematical formulation.
- Java programming.
- Research methodology.
- Problem solving.
Key words: Cellular Automata – Reaction-diffusion – Multi-agent Systems – Clustering.