This paper introduces a technique for solving the flowshop scheduling problem. The major idea is to partition the set of feasible solutions into regions in order to diversify the search that is used on a Genetic Algorithm variation. The population is formed by every distinct subject and it is carried out constructively in such a way that any iteration guarantees a diversification on the search for a feasible solution. The problem is a very well known NP-Hard problem and it imposes great challenges for determining its optimal solution in the practice. Computational experiments are reported for the literature instances and the obtained results are compared with other techniques.

Scheduling Problem, Genetic Algorithm, Permutations.