Input-output Linearization of a Pilot Pressure Plant using a Dynamic Neural Network Model

V.M. Becerra, F. Garces (UK), J.M.F. Calado, and P.M. Silva (Portugal)


Nonlinear control theory, feedback linearization, dynamic neural networks


This paper describes a method for feedback linearization based on dynamic neural network (DNN) models. The pa rameters of these models can be identified to reproduce the behaviour of a dynamic system. The model is then used to compute the input-output linearization transformations. The same model is used as an open loop observer to provide state information to compute the input-output linearization. An external linear controller is used to control the input output linearized system. The method is applied to a pilot scale pressure plant.

Important Links:

Go Back