Article ID: | iaor1995545 |
Country: | United Kingdom |
Volume: | 32 |
Issue: | 8 |
Start Page Number: | 1759 |
End Page Number: | 1773 |
Publication Date: | Aug 1994 |
Journal: | International Journal of Production Research |
Authors: | Yeo K.T., Sim S.K., Lee W.H. |
Keywords: | artificial intelligence: expert systems, neural networks |
The objective of this research is to apply the neural network approach to the dynamic job shop scheduling problem. A feed-forward back propagation neural network is designed and trained to recognize the individual contributions of traditional dispatch rules. The network is incorporated into an expert system which activates the network according to the prevailing shop environment. The effectiveness of the approach is compared with the traditional dispatch rule approach as well as a composite rule expert system. Results of scheduling with a neural network show that the network is able to perform well against its component factors for job lots with varying arrival rates.