Population-based neighborhood search for job shop scheduling with interval processing time

This paper applies interval number theory to production scheduling for its advantage in uncertainty modeling. A job shop scheduling problem with interval processing time is first described and then a population-based neighborhood search (PNS) is presented to optimize the interval makespan of the problem. In PNS, an ordered operation-based representation is used and a decoding procedure is constructed by using operations of interval numbers, in which there are no approximate treatments. It is proved that the possible actual makespan of each schedule are contained in its interval makespan. A swap operation and binary tournament selection are applied to update the population. PNS is finally tested by using some instances and computational results show that PNS can provide better results than some methods from the literature.

► Interval number theory is applied to production scheduling for its advantage in uncertainty modeling.
► A population-based neighborhood search (PNS) is used to optimize the interval makespan of the problem.
► It is proved that the possible actual makespan of each schedule are contained in its interval makespan.
► In PNS, an ordered operation-based representation, a swap and tournament selection is used.
► Computational results show that PNS can provide better results than methods from the literature.