Modelling mono-product production systems by means of a neural network: continuous and discrete systems, represented by continuous models

This paper presents an approach of modelling and optimizing stages which intervene in the production systems designing or running process. The submitted modelling it; based on the utilization of neural networks, for the operative parts as well as for the control parts. The point is to reinforce the graphical approach through the implementation of generic elements in the form of neural networks for the various entities of a production system. The interest of this approach is two-fold. It allows us to make full use of the extended simulation, optimization and control possibilities offered by neural networks. Also, it contributes in avoiding risks inherent in the writing and handling of equations. An application using most of the developed elementary models concludes the paper.