Title :
Non-energy based neural networks for job-shop scheduling
Author :
Jeng, MuDer ; Chang, ChunYu
Author_Institution :
Nat. Taiwan Ocean Univ., China
fDate :
2/27/1997 12:00:00 AM
Abstract :
A synchronous neural network architecture that implements a heuristic rule is proposed for solving the job-shop scheduling problem. The proposed rule can obtain better near-optimal solutions than some commonly used heuristic rules. The approach resolves drawbacks in prior work based on energy functions such as invalid solutions, local minima and sensitivity to initial inputs
Keywords :
neural nets; optimisation; resource allocation; scheduling; energy functions; heuristic rule; initial inputs; job-shop scheduling; local minima; near-optimal solutions; nonenergy based neural networks; sensitivity; synchronous neural network architecture;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19970269
