Analytical study of Wind power prediction system by using Feed Forward Neural Network

The power produced through the wind changes very quickly because of the constant fluctuation of the speed of wind and direction. This is essential for the power industry to have the ability to forecast the Power generated by the wind for the power control and management. A short term method of forecasting the wind power of a wind power plant (WPP) is introduced in this paper which is trained by Artificial Neural Network (ANN) which had been used for many years for optimizing the results of various problems in various sectors. There are two stages used for wind-power prediction; in the first stage the data of generated wind-power, wind pressure, wind direction & wind speed were collected from Jodhpur, Rajasthan in the west India. In the second stage a model was prepared to predict wind-power by applying the Levenberg-Marquardt optimization approach which was utilized as it aids the finest training rate pursued as a back propagation algorithm for the multilayer Feed Forward ANN model using MATLAB
^®
R2013 ANN toolbox. The implementation of the network architecture, training of the Neural Network and simulation of final conclusions were all fruitful along an extremely huge level of efficiency concluding into hourly AC system output. The predicted result was accurate with respect to the measured data.