First Generation Multi-agent Models and Their Upgrades

Multi-agent systems consist of interactive and independent agents of different kinds in a "world" of the computers. The key issue of multi-agent modelling is its ability to produce emergent phenomena at macro level from "micro-behaviour". For now this approach became a widely used methodology in socio-economics and ecology. This paper presents three famous first generation models and then drafts some of their upgrades, especially the agent-based computational economics, the spatial planning approach and the ecological models. Finally some conceptual developments are presented and discussed. Classification-ACM-1998: I.2.11; Distributed Artificial Intelligence, Multiagent systemssimulation, early multi-agent models