A self-adaptive multi-agent system consists of agents. Each agent is autonomous and can make decisions. The self-adaptive system can adapt according to the requirements and surrounding environment.
When I was working at CSIRO Marine Research, we wrote large (regional scale) simulation models of human/ecosystem interactions. Many of the entities in the simulations made decisions based on their internal state, the state of their local environment, and the constraints which they had to abide by.
A simpler model was the PMEZ model which looked at the potential for contaminants to percolate through the food chain. In this model, the behaviour of an organism was conditioned by its need to migrate (or not), how hungry it was, the local water temperature and depth, and the time of day. Whether an individual fish ventured into "unpleasant" water to forage was based on how desparate it was for food, and (to a degree) how likely it was to find it.
Book NWSJEMS Ecosystem model specification with an agent based framework
- Badges
- Science method