Neural Network Computing for the Electric Power Industry

F. Control

Neural Control Systems

Bernard Widrow and Françoise Beaufays^*

Department of Electrical Engineering

Stanford University

Stanford, CA 94305-4055

1 Introduction

In the past decade, the field of neural networks has grown from a mere handful of researchers to thousands spread throughout academics and industry worldwide. Applications have been developed in domains as varied as character recognition, pattern classification, dynamic system control, and stock market prediction. In this paper, we concentrate on the application of feedforward multilayer neural networks to nonlinear control. After a brief introduction on adaptive linear combiners and multilayer neural networks, we present one of the most famous algorithms used in neural control: ...