Power System Load Modeling using a RBF GRNN with Self-starting Centers

G.L. Lebby, K.M. Stevenson, and G.H. Shi (USA)


Artificial Intelligence, Power System, Load Modeling, Neural Networks, GRNN, Harmonics.


A Power System Load (PSL) model of a southeastern electric power cooperate 1 has been developed that uti lizes heuristical and analytical methods for establishing the hourly and daily base load, growth rate, seasonal weather load, and a non-seasonal (random) weather load compo nents. Load models are tools, which have a wide range of application in the electric power industry. These uses in clude monitoring load management policies, helping with the generator commitment problem by providing short term forecasts, and aiding system planning byInstitute of Electrical and Electronics Engineers (IEEE) Association for Computing Machinery ( term forecasts. An innovative method based upon analyz ing four years of hourly data is proposed to generate an op timum model structure from a set of preselected heuristic basis functions. The result is a PSL capable of providing high quality short term forecasts.

Important Links:

Go Back