Abstract:
MRP technique is widely used in manufacturing planning system. A weak point
of MRP system is that it assumes fixed lead-time and infinite capacity of machines.
Therefore the generated production plan is infeasible. This research aims to propose the
MRP system with capacity constraint (MRPC). The proposed system has two main
steps. Firstly, the proposed system generates a sequence of operations for all work
centers. There are two types of sequences generated by this step. The first one is to
apply the same sequence for all work centers as permutation concept called MRPCP.
The second one is different from the first one in that the sequences of operations on
work centers are not necessary to be the same and called as MRPCNP. Secondly,
a linear programming is applied to determine the optimal start times of operations on
each work center. Five key performance measures are considered namely, total
tardiness, number of tardy orders, total earliness, number of early orders, and average
flow-time. The performance measures of MRPC are compared to those of the MRP and
the finite capacity MRP (FCMRP) system. The results show that the MRPC and FCMRP
system are more flexible than the MRP system since the planner can generate the
desire performance by changing weights in the objective function. The total tardiness
and average flow-time obtained from MRPCP system is the best compared to that of
the MRPCNP and FCMRP system. The FCMRP system obtains the lowest earliness.
Furthermore, the dispatching rule named MST outperforms the others for all
performance measures. Based on the overall performance measures, the MRPCP
outperforms MRPCNP and FCMRP systems.