The Simulation Of Ill-Defined Processing Operations Using Connectionist Networks

This paper proposes a new methodology to model ill- defined processing operations using neural nets (NN's). A process with many variables (large dimensionality) cannot be modelled adequately if limited process data are available. This problem of multi-dimensionality is addressed and an approach suggested to reduce the dimensionality using NN's. The result of this modelling methodology is a combined model of phenomenological expressions and NN's. The model can successfully be incorporated in a dynamic simulator of the process. A new approach to conduct modelling on the basis of continuous data collected directly from an industrial processing unit, is also proposed. The modelling methodology is applied to a typical processing operation, the resin-in-pulp (RIP) process.