Job Shop Flow Time Prediction using Neural Networks

Abstract

In this paper we investigate the use of Artificial Neural Networks (ANN) for flow time prediction and, consequently, to estimate due dates (DD) in a hypothetical dynamic job-shop. The effectiveness of the proposed ANN based DD assignment model is evaluated comparing it performance with the performance of two dynamic DD assignment rules proposed in the literature: Dynamic Total Work Content, and Dynamic Processing Plus Waiting. Results show that ANN based DD assignment models are more effective than, not only available static DD assignment rules, as concluded by other researchers, but also than the more effective Dynamic DD assignment rules.