Neural networks as complementary direct controller of nonlinear plants

A neural network complementary controller for fine tuning performance of the approximate controllers of nonlinear plants is suggested. This approach of using neural networks in control allows placing the maximum amount of knowledge about a plant in its controller design, while leaving the final fine tuning to the neural networks. This method does not put too much restriction on the type of plant to be controlled. The system designed by this method has stable performance for the type of inputs for which it has been trained. This approach does not require the knowledge of the appropriate form of controller output for each given input and does not require identification of the plant or its inverse model. The similarities and differences of this approach and other methodologies are explained.<
>