Neural network computational models offer significant promise for improving several aspects of power plant operation. This paper provides a brief but comprehensive review of neural networks. Various learning paradigms are discussed, as is the use of neural networks for solving optimization problems. Several applications relevant to the utility industry are described briefly. These include load forecasting, security monitoring, turbine backpressure optimization and process modeling. These are applications in which the ability of neural networks to learn complex mappings from training data is used to develop estimators or predictors of various properties of interest. Characteristics that a problem should possess for a neural network approach to be viable are also discussed.
|Original language||English (US)|
|Number of pages||6|
|Journal||Proceedings of the American Power Conference|
|State||Published - Dec 1 1990|