The implementation of neural networks for the optimization of the production scheduling

The work presents the application of a constraint satisfaction adaptive neural network to job-shop the scheduling problem. The main idea of the CSANN method is described. In particular, the capacity of the net for adaptation to constraints of a specific problem is presented. A computer experiment is conducted to find the Johnson criterion (the minimal total time of the performance of all operations). The criterion is mainly found as a function of the number of iterations of the computing process. Achieved results are compared with the genetic algorithm AGHAR worked out for the solving of such problems.