Constraint satisfaction algorithms: edition of timetables in the license-master-doctorate system

Abstract

In this paper, we studied some algorithms for solving constraint satisfaction problem (CSP) and then applied them to solve the problem of generating schedules in a university setting. In other words, we studied the genetic algorithm, the simulated annealing, the hill climbing, a hybridization of the genetic algorithm and the simulated annealing as well as a hybridization of the genetic algorithm and the hill climbing. These algorithms have been tested on the problem of scheduling in a university environment. The hybrid uses hill climbing or simulated annealing to improve each individual in the starting population to a certain stopping point. These individuals are then sent to the genetic algorithm. Our results show that the hybridization of the genetic algorithm with a metaheuristic gives better execution time and performs better as the problem size increases compared to the classical genetic algorithm.