A study of group size and communication in an evolving fuzzy-controlled population

This research effort is an investigation into communication activity in a distributed set of software agents. Agents exist in a predator-prey environment. Movements of prey agents are evolved upon a Mamdani type fuzzy inference system. Probabilistic predation and starvation forces, along with simulated communication activity act upon agents, causing them to cluster. Examined here is the correlation between mean cluster size after the population has sufficiently evolved, versus the average agent communication activity. Communication activity creates a loyalty to remain in a cluster for security rather than to change cluster membership. A mean r² correlation of 0.87 was observed with this system, thus lending credence to our hypothesis.