| Popis: |
Wind power is dependent on several factors, and its prediction by complex mathematical model makes it prone to errors. Artificial neural network-based prediction models provides a solution to this problem. In this work, artificial neural network is employed for the energy prediction using wind direction and wind speed. Though the power curve is based on factors like wind velocity, air density, and swept area, there are many other dynamics which affect the performance of the wind turbines. A mathematical model is generated in the present work for wind energy production. The data is taken and fed to the back propagation algorithm in neural networks. The algorithm is run for thousand iterations, and by adjusting the weights, the model is created. It is observed that the prediction almost follow the actual power curve and hence can be used to predict different data of speed and direction. |
