Evaluation of wind energy production using weibull distribution and artificial neural networks

Abstract

Wind turbine power generation planning requires production estimation. Wind power is uncertain depending on the location, wind speed and wind turbine efficiency. This paper presents a method for evaluating wind energy production using Weibull distribution and Artificial neural networks to compare the data recorded by Promthep Alternative Energy Station, Phuket, Thailand. The results show that wind energy estimation using artificial neural networks produces the most accurate results. Mean Absolute Percentage Error is used to determine the minimum error value. Minimal error of training data is 2.524% and the test data is 3.3041%.