Applying GENET to the JSSCSOP

Abstract

GENET is a local search approach with a neural network connectionist architecture for solving constraint satisfaction problems by iterative improvement and incorporates a learning strategy to escape local minima. In this paper, a method within the framework of propagation of posted new constraints and based on the progressive stochastic search of GENET for solving the job shop scheduling constraint satisfaction optimization problem (JSSCSOP) will be presented. The experimental results show that the performance of our method gets competitive when the domain of each variable is not big, even if the size of the problem instances increases. © Springer-Verlag 2004.