Neural networks for job-shop scheduling

A neural network structure has been developed which is capable of solving deterministic job-shop scheduling problems, part of the large class of np-complete problems. The problem was translated in an integer linear programming format which facilatated translation in an adequate neural network structure. Use of the presented structure eliminated the need for integer adjustments. Elementary precalculation is performed with the objective to reduce the search space allowing more rapid calculation of feasible solutions. In this precalculation the earliest possible starting times of the operations are calculated and set as tresholds in the network. The neural network structure was reliable in simulated operation and its performance was superior to structures which have been presented previously. The network structure always produces feasible solutions, in less time, without the application of integer adjustments.