Abstract: Reconfigurable manufacturing systems (RMS) are manufacturing systems designed for part family, which have both customized flexibility and high productivity. RMS adapt to the fast market fluctuation by reconfigurating. The design objective of RMS is to find the optimal and sub-optimal cost configuration paths across the lifetime of manufacturing systems. First, the cost model for every production period of RMS, the reconfigurating cost model and the lifetime cost model of RMS are constructed. Second, the directed graphs of configurations of RMS at every production period are constructed. Finally, based on the directed graphs, the optimal and K–1 sub-optimal cost configurations of RMS at every production period are found out by using Dijkstra algorithm and double-sweep algori- thm. According to the optimal and K–1 sub-optimal cost configurations of RMS at every production period, the reconfi- gurating cost model and the lifetime cost model of RMS achieved above, the reconfigurating costs between every configuration at the up-production period and every one at the down-production period are calculated, then the econfigurating policy directed graph across the lifetime of RMS is constructed. Using the Dij-kstra algorithm and double-sweep algorithm again, the optimal and K–1 sub-optimal reconfigurating policy across the lifetime of RMS is figured out. Finally, a case study illustrates the deve-loped method is valid.

Key words: Dijkstra algorithm Double-sweep algorithm, Reconfigurable manufacturing systems Reconfigurating policy, Reconfigurating cost Graph theory