Variable neighborhood search algorithms for the permutation flowshop scheduling problem with the preventive maintenance

This paper addresses to the permutation flowshop scheduling problem while considering the preventive maintenance in the non-resumable case. The criterion to be optimized is the makespan. Two variable neighborhood search algorithms are proposed. In the first algorithm, only one initial solution is generated according to a constructive heuristic. In the second algorithm, a learning process using a probabilistic model is introduced to the variable neighborhood algorithm in order to generate the initial solution. The computational results show the high performance of the proposed algorithms according to the compared approaches. Besides, the change of the initial solution during the optimization procedure may improve the performance of the variable neighborhood search algorithm.