Novel neural internal model control structure

This paper investigates the application of neural networks to the modelling and control of nonlinear systems. Neural network based plant modelling is discussed first with a powerful parallel BFGS based training algorithm proposed for the rapid off-line training of such models from plant data. A novel nonlinear Internal Model Control (IMC) strategy is suggested, that utilises a nonlinear neural model of the plant to generate parameter estimates over the nonlinear operating region for an adaptive linear internal model, without the problems associated with recursive parameter identification algorithms. Unlike other neural IMC approaches the linear control law can then be readily designed. A Continuously Stirred Tank Reactor, (CSTR), was chosen as a nonlinear case-study for the techniques discussed in this paper.