Author(s)

K. Yamazaki, S. Kundu & M. Hamano

Abstract

Genetic based learning and adaptation algorithms called Genetic Programming (GP) is used to derive optimal control rules for adaptive structures or the variable geometry (VG) structures. Autonomous distributed control rules which can evolve are proposed, as opposed to non- autonomous, centrally controlled and predetermined control rules. These rules, which have tree like hierarchical structures embedded in them, form an individual in the GP population. Optimal rules are searched for by the application of randomized genetic operations on the tree structures of these rules. In the rules, the main tree contains condition trees and rule selection trees as its function nodes and control trees as its

Keywords