The Ant Colony Metaphor for Searching Continuous Design Spaces

This paper describes a form of dynamical computational system--4he ant colony---and presents an ant colony model for continuous space optimisation problems. The ant colony metaphor is applied to a real world heavily constrained engineering design problem. It is capable of accelerating the search process and finding acceptable solutions which otherwise could not be discovered by a GA. By integrating the Pareto optimality concept within the selection mechanism in GAs and Ant Colony it is possible to treat both hard and soft constraints. Hard constraints participate in a penalty term while soft constraints become part of a multi-criteria formulation of the problem.

• Type: Conference
• Year: 1995
• Where: AISB
• Authors: George Bilchev, Ian C. Parmee
• Keywords: AISB 1995, Ant Colony, Ant Colony Metaphor, Ant Colony Model, Artificial Intelligence