Genetic algorithm for some partitioning and sequencing problems

In this study, a generic optimization problem Part-Seq is formulated. It consists in partitioning the set of objects into smaller subsets and finding optimal sequences of elements in each subset. One well-known application of this problem is the vehicle routing problem. Another application considered in this paper is a multi-product sheduling problem that arises in chemical industry. A genetic algorithm for solving the problem is developed. The desing of the algorithm is quite generic and requires small adaptation to a particular application. To improve the solutions in the genetic algorithm, two additional steps are introduced: the parallel dynamic programming algorithm implemented on a graphical processing unit and a pool of best found permutations. The computational results are presented.