An enhanced loading model for the probabilistic workload control under workload imbalance

Abstract

Load-oriented manufacturing control (LOMC), a well known probabilistic approach to workload control, is based on limiting and smoothing workload using one static parameter for each workcentre, called load limit (LL). The value of this parameter is set by the shop managers based on the planned lead time at each workcentre. In this paper the use of LL is shown to be inappropriate for the smoothing of workloads when the workload is not sufficiently balanced. We propose to enhance the LOMC model by introducing two sets of parameters:

(i) limiting parameters (LPs), that are statical parameters of the workcentres, set by the shop managers. LPs are used to limit the workload released to the shop;

(ii) smoothing parameters (SPs), that are dynamical parameters of the workcentres, computed as a function of their real workload. SPs are used to smooth the jobs workload over downstream workcentres.

A simulation model was used to compare the enhanced model, based on two parameters sets, with the traditional LOMC model, based on a single parameter set. The simulation runs were earned out with different conditions of due-date assignments, dispatching rules and production mix. The statistical analysis performed on experimental results confirmed that the enhanced model achieves significantly better due dates under unbalanced workload conditions.

Keywords:

• loading
• workload control
• job shop