We propose a new approach towards Particle Swarm Optimization named Agent-based PSO. The swarm is elevated to the status of a multi-agent system by giving the particles more autonomy, an asynchronous execution, and superior learning capabilities. The problem space is modeled as an environment which forms clusters of points that are known to be non-optimal and this transforms the environment into a more dynamic and informative resource.
Ahmad, Raheel, Lee, Yung-Chuan, Rahimi, Shahram and Gupta, Bidyut. "A Multi-Agent Based Approach for Particle Swarm Optimization." (May 2007).