Abstract :
Artificial neural networks have been applied to power systems in such areas as control, load forecasting, monitoring, diagnosis, and analysis. Their prevalence in the literature and product lines of the electric utility industry has almost overworked the descriptor “intelligent”. However, ANNs do not have an IQ. They exhibit attributes that humans often associate with intelligence: an ability to learn, to communicate, to perceive, or to reason. The objective of this tutorial is to give a basic understanding of ANNs and computational intelligence. This is accomplished by discussing the historical and biological basis of ANNs and reviewing two representative architectures from a simple taxonomy for ANNs. In this tutorial, the use of the term architecture is used to imply both a network topology and a learning rule. Resources for further studies and free software are also recommended
Keywords :
learning (artificial intelligence); load forecasting; neural net architecture; neural nets; power system analysis computing; power system control; artificial neural networks; biological basis; computational intelligence; diagnosis; electric utility industry; learning ability; learning rule; load forecasting; monitoring; network topology; neural net architectures; power system analysis; power system control; power systems; self-organisation; supervised learning; unsupervised learning; Artificial neural networks; Computational intelligence; Computer architecture; Control systems; Load forecasting; Monitoring; Power industry; Power system analysis computing; Power system control; Power systems;
