The application of multiparadigm simulation techniques to manufacturing processes

The simulation of manufacturing processes and manufacturing systems using c omputers has been carried out for several decades. The use of discrete even t simulation models for production systems, material planning, machine grou ping and various other key problems of resource allocation have been well r esearched and a number of ready to use commercial products are available. I n recent years, the use of artificial intelligence simulation and modelling techniques for certain manufacturing processes have led to significant adv antages over conventional techniques. However, the benefits of using AI tec hniques have not been fully exploited in existing simulation systems. Owing to the different way that some AI techniques process or represent simulati on data compared to conventional mathematically based methods, it is not al ways easy to include AI techniques in existing simulation systems with thei r inherent structures for representing simulation data. in this paper, a co ncept which allows the use of AI techniques such as artificial neural netwo rks, genetic algorithms and fuzzy sets as well as traditional mathematical techniques for the simulation and modelling of certain manufacturing proces ses is presented and discussed. Methods for the exchange of simulation data are demonstrated using the milling process as an example. Furthermore, the notion of distributed simulation and the determination of essential proces s parameters by remote access to common simulation databases using a client /server model has been reviewed.