A Multi Layered Perceptron Networkl for Wind Speed Generation

P.R.J. Campbell and K. Adamson (UK)

Keywords

Neural Networks, Perceptron, Backpropagation, WindEnergy

Abstract

As wind energy continues to increase in popularity as a viable alternative to fossil fuels it is important that the wind industry has an efficient and effective method for site selection. The selection of sites is based mainly on the locations wind regime. Current methods of site assessment are based on a minimum of one year's recorded wind speeds. This paper presents a multi layered perceptron network for the generation of wind speed. The presented 4 layer model which predicts wind speed to a high degree of accuracy, and shows significant scope for application as either a site selection tool or as a method for filling gaps within a dataset.

Important Links:



Go Back


IASTED
Rotating Call For Paper Image