Computationally efficient neural predictive control based on a feedforward architecture

A new strategy for integrating system identification and predictive control is proposed. A novel feedforward neural network architecture is developed that allows the nonlinearity to be mapped onto a linear time varying term. The linear time varying neural network model is augmented with a Kalman filter to provide unmeasured disturbance estimation and compensation for model uncertainty. The structure of the model lends itself naturally to a neural predictive control formulation. The computational requirement of this strategy is significantly lower than that of optimization of the nonlinear neural network model, with comparable control performance, as illustrated on a challenging nonlinear chemical reactor with input multiplicity