The paper presents the application of the constraints satisfaction adaptive neural network with heuristics algorithms to various cases of the job-shop scheduling problem. The paper describes the main idea of the CSANN method and three possible structures of the CSANN network designed for three classes of production cases. CSANN itself finds and optimizes feasible solutions of the production schedules considering the sequence and resource constraints. CSANN network works together with three heuristics algorithms: the first one eliminates dead lock situations, the second one eliminates empty periods of time without any operation on any machine, the third one obtains a new set of the initial value needed for iterations. The structure of CSANN has been adopted for a class of production cases. The first structure has been designed for the solving serial flow production case without groups of technologically changeable machines and has been originated from literature. The second structure has been created by authors and it is able to solve serial production cases with groups of technologically changeable machines. The third structure also has been built by authors and is able to solve parallel production cases without groups of technologically changeable machines. The paper presents results of computer experiments proceeded for second and third class of production cases. The results of experiment achieved for the serial flow production case with groups of technologically changeable machines are compared with results achieved by other method - genetic algorithm AGHAR. Authors point in the paper that it seems to be possible to develop the CSANN further so the optimization of more complicated processes will be possible. As a next step the structure which is able to optimize the parallel process with groups of technologically changeable machines will be constructed.
[1]
Shengxiang Yang,et al.
Constraint satisfaction adaptive neural network and heuristics combined approaches for generalized job-shop scheduling
,
2000,
IEEE Trans. Neural Networks Learn. Syst..
[2]
J. J. Hopfield,et al.
“Neural” computation of decisions in optimization problems
,
1985,
Biological Cybernetics.
[3]
Leon O. Chua,et al.
Neural networks for nonlinear programming
,
1988
.
[4]
John J. Hopfield,et al.
Simple 'neural' optimization networks: An A/D converter, signal decision circuit, and a linear programming circuit
,
1986
.
[5]
P. Antczak,et al.
The implementation of neural networks for the optimization of the production scheduling
,
2004,
2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).