Genetic algorithms based on primal-dual chromosomes for royal road functions

Abstract

Genetic algorithms (GAs) have been broadly studied by a huge amount of researchers and there are many variations developed based on Holland’s simple genetic algorithm (SGA). Inspired by the idea of diploid genotype and dominance mechanisms that broadly exists in nature, we propose a primal-dual genetic algorithm (PDGA). PDGA operates on a pair of chromosomes that are primal-dual to each other in the sense of Hamming distance in genotype. We compare the performance of PDGA over SGA based on the Royal Road functions, which are specially designed for testing GA's performance. The experiment results show that PDGA outperforms SGA on the Royal Road functions for different performance measures.