Particle swarm optimization

This paper introduces particle swarm optimization, a method for optimizing nonlinear functions inspired by the social behavior of swarms such as flocks of birds or schools of fish. The authors outline the evolution of several related paradigm variants of the method and describe the implementation details of one particular version chosen for evaluation.

The chosen paradigm is subjected to benchmark testing, with the authors proposing applications to nonlinear function optimization and to the training of neural networks. The paper also draws out the relationships between particle swarm optimization and both artificial life research and genetic algorithms, positioning it as a distinct population-based optimization approach within that broader landscape.