Zooming In, Zooming Out: A Framework for Hierarchical Genetic Algorithms

摘要

We present a framework of algorithms and techniques involving hierarchical genetic algorithms. These algorithms attempt to generate models and solutions at both the structural and atomic levels simultaneously using atomic and global chromosomes. Complex adaptive emergent systems (CAESs) are comprised of heterogeneous, interacting, adaptive agents that exhibit the properties of self-organization, emergence, and connectivity [Ahmed2005J. This paper presents early work and is meant to serve as a starting point in the synthesis of complex adaptive emergent systems using HGAs. Two implementations are discussed1 that exemplify the compositional nature of hierarchical genetic algorithms and their applicability to both modeling and iteratively generating complex adaptive emergent systems.