Nonlinear plant control using neural networks: a design procedure through linearization

Abstract

A three stage procedure to design neural net controller for nonlinear plants is developed. The design is based upon linearization of the plant at each operating point. In stage one, the input-output data of the plant is used to train a neural net model of the plant. In stage two, the plant is controlled using a time varying linear controller based on the linearized plant model at each operating point. The neural net model of the plant identified in stage one is used to obtain the linearized models. In stage three, a neural net is trained to replace the time varying linear controller. Effect of noise has also been investigated. Simulation results of SISO as well as MIMO plants have been presented.