Minimizing total flowtime and maximum earliness on a single machine using multiple measures of fitness

摘要

The intent of this research is to investigate methods to use genetic algorithms to find the set of efficient solutions to a bi-criteria problem. We propose a general methodology which is characterized by using different criteria upon which the decision to retain chromosomes into the next generation is made. We perform elite reproduction based on two general measures of "eliteness": non-dominated in the current population and performance measured in terms of each criterion individually. We investigate its performance on a specific bi-criteria scheduling problem, minimizing total flowtime and maximum earliness on a single machine.