A mixed integer program for cyclic scheduling of flexible flow lines

A new mixed integer programming formulation is presented for cyclic scheduling in flow lines with parallel machines and finite in-process buffers, where a Minimal Part Set (MPS) in the same proportion as the overall production target is repetitively scheduled. The cycle of parts in an MPS is not determined a priori, but is obtained along with the optimal schedule for all parts. In addition to the cyclic scheduling, a cyclic-batch scheduling mode is introduced, where within the MPS the parts of one type are processed consecutively. Numerical examples are included and some results of computational experiments are reported.