A Testbed for Solving Optimization Problems Using Cultural Algorithms.
Chan‐Jin Chung, Robert G. Reynolds · 1996
This paper introduce a cultural algorithm based testbed which allows one to plug and play various combinations of evolution components for solving constrained numerical optimization. Our cultural algorithm framework combines weak search method with knowledge representation scheme for collecting and reasoning knowledge about individual experience. Currently genetic algorithm based software package GENOCOP(GEnetic algorithm for Numerical Optimization for COnstrained Problems) and rudimentary EP(Evolutionary Programming) are embedded in the cultural algorithm framework. Preliminary results suggest that the belief space is an important contributor to the problem solving process for both systems when the number of constraints on the problem become large enough. 1.0 Introduction Recently much attention has been paid to the constrained optimization problems using evolutionary computation techniques in order to solve large scale real-world problems [Michal91,94a,95a][Fogel95a] The formal defi...