A genetic algorithm with adaptable parameters

An approach to the self-adaptation of the parameters of a genetic algorithm to the optimization problem is described. The characterizing feature of the system presented is the structure of the evolving individuals, that are more similar to cells rather than to isolated genes with the introduction of a program chromosome besides the chromosome of the optimization problem variables. The basic scheme of the evolutionary optimizer is a multiple objective genetic algorithm based on the introduction of the dominance criteria in the selection mechanism. Some preliminary results are presented on a multimodal single-objective test function and the directions of future work are outlined.