A decomposition-based approach to flexible flow shop scheduling under stochastic setup times

Research on production scheduling under uncertainty has recently received much attention. This paper presents a novel decomposition-based approach (DBA) to flexible flow shop (FFS) scheduling under stochastic setup times. In comparison with traditional methods using a single approach, the proposed DBA combines and takes advantage of two different approaches, namely the Genetic Algorithm (GA) and the Shortest Processing Time Algorithm (SPT), to deal with uncertainty. A neighbouring K-means clustering algorithm is developed to firstly decompose an FFS into an appropriate number of machine clusters. A back propagation network (BPN) is then adopted to assign either GA or SPT to generate a sub-schedule for each machine cluster. Finally, an overall schedule is generated by integrating the sub-schedules of the machine clusters. Computation results reveal that the DBA is superior to SPT and GA alone for FFS scheduling under stochastic setup times.