Staff Scheduling By A Genetic Algorithm

This paper describes a genetic algorithms approach to amanpower-scheduling problem arising at a petrochemical company. althoughgenetic algorithms have been successfully used for similar problemsin the past, they always had to overcome the limitations of theclassical genetic algorithms paradigm in handling the conflict betweenobjectives and constraints. the approach taken here is to use an indirectcoding based on permutations of the personnel’s, and a heuristicdecoder that builds schedules from these permutations. computationalexperiments based on 52 weeks of live data are used to evaluate three differentdecoders with varying levels of intelligence, and four well-knowncrossover operators. the results reveal that the proposed algorithm isable to find high quality solutions and is both faster and more flexiblethan a recently published taboo search approach