Evolutionary exploration of dynamic swarm behaviour

Abstract- Dynamic swarm behaviour is difficult to design manually. Multiple, simultaneous interactions among a large number of agents, make the relationship between a parameter change and the corresponding change in global behaviour not obvious. This paper presents breeding experiments of dynamic swarm behaviour patterns using an interactive evolutionary algorithm. Specifically, eight scalar parameters that influence swarm behaviour dynamics in a 3D swarm simulation are interactively evolved to produce agents that collectively fly in line, ring, and figure-eight formations.

• Type: Conference
• Year: 2003
• Where: CEC
• Authors: Henry Kwong, Christian Jacob
• Keywords: Artificial Intelligence, CEC 2003, Dynamic Swarm Behaviour, Swarm Behaviour, Swarm Behaviour Patterns