Neural-net-based direct self-tuning control of nonlinear plants

Use of neural networks for direct self-tuning control of stochastic nonlinear plants has been proposed. The control is based upon inverse modelling of a pseudo-plant. The input to the pseudo-plant is same as the plant input while its output consists of a linear combination of the plant input and output. The controller is directly identified as a mean square optimal inverse estimator of the pseudo-plant. This approach allows the control of inverse unstable plants. Local convergence properties as well as results of simulation studies are presented.