NEURAL MODELING OF CHEMICAL-PLANT USING MLP AND B-SPLINE NETWORKS

This paper demonstrates the potential of the B-spline neural network ( BSNN) for the modelling of nonlinear processes. The performance of thi s paradigm is compared with the multi-layer perceptron (MLP), for the modelling of experimental and industrial case-studies. The experimenta l pH neutralisation plant of the University of California at Santa Bar bara (UCSB) is modelled using both networks. This case-study demonstra tes the higher performance of the B-spline network, for rapid on-line model adaptation. The second case-study focuses on the prediction of v iscosity for an industrial polymerisation reactor. Predictive models o f viscosity are developed, based on both networks to predict over and hence remove the measurement time-delay introduced at the viscometer. A novel disturbance modelling approach is also developed here, and dem onstrated to perform excellently in on-line tests. Copyright (C) 1997 Elsevier Science Ltd.