Combining constraint programming and genetic algorithm for dynamic scheduling problems

Abstract

This paper introduces a new method based on constraint programming (CP) and genetic algorithm (GA) for solving dynamic scheduling problems. The proposed approach allows us to handle scheduling problems with large sizes (i.e. search spaces are too large). Our idea is to break up the search space into disjoined sub-spaces by the genetic algorithm. To each individual of the population is associated a sub-space. Each sub-space is represented by a sub-CSP which is easier to solve than the original scheduling problem. Our first experimentations are addressed to the Endoscopy Unit scheduling in dynamic way.