Author : Wen-tsao Wang
Publisher :
Page : 366 pages
File Size : 21,66 MB
Release : 1989
Category : Order picking systems
ISBN :
Book Description
Automated Storage and Retrieval (AS/R) Systems have had a considerable impact on manufacturing and distribution processes. The configuration of AS/R systems vary considerably, depending on the particular application. The effect of different retrieval scheduling rules is analyzed and compared for three different unit-load AS/R systems operating under dual-command cycle. The AS/R systems differ in the number of docks (or input/ output points) per aisle. The system is based on that there are two storage (S1, S2) and two retrieval (R1, R2) sources. The three AS/R systems are: (1) one dock per aisle handling both storage and both retrievals; (2) two docks, one at each end of the aisle, with storage requests from S1 and S2 retrieved by either RI or R2; and (3) two docks, one at each end of the aisle, with storage requests from Si retrieved by RI and storage requests from S2 retrieved by R2. The same scheduling rule, first- come - first serve /closest-open-location, was used for storage. However, retrieval scheduling rules analyzed in the study were first-come-first-serve, nearest-neighbor, and nearest-neighbor with a maximum wait time limit. A simulation model was developed for each combination of scheduling rule and layout and three performance measures were examined--throughput, mean waiting time, and maximum waiting time. The results obtained from this investigation show that the efficiency of the AS/R system can be improved oy the introduction of two docks in each aisle when the input pallets are stored in the closest-open-location, and the input/output pallets for the two docks are independent of each other (layout 3). In general, each performance measure reflects an independent optimum solution for each layout, or combination of layout and scheduling policies. The use of one of these independent solutions can serve to optimize selected AS/R configurations with regard to the given performance measures, but at the same time other performance measures may be affected adversely. Thus, there is no one global optima; the optimum is a function of the performance measure used.