Research On Short-term Prediction Methods Of Wind Speed And Wind Turbine Power Generation

Short-term prediction of wind speed and wind turbine power generation on wind farm is important for utility of wind power.Based on accurate prediction of wind speed and power generation,the wind generating plan can be efficiently accommodated to mitigate the impaction from instable wind power on wind grids.At present there are two main methods for short-term wind speed prediction,one is based on Numerical Weather Prediction,and another is based on statistical methods.This paper analysis the properties of wind speed with real measured data obtained from a wind farm form 2006.3 to 2007.2 in Zhangbei area of China. The sampling periodic of the wind speed data is 10 minutes.And also introduce several wind speed models.Then Time series and Artificial Neural Network are adopted to predict the wind speed.Meanwhile,to deal with the defects(low convergence rate and easy to fall into local minimum value) of Back-Propagation Neural Network(BPNN),this paper uses momentum BP algorithm to speed up the convergence rate,and takes the advantage of the genetic algorithm(GA) to optimize the structure,weights and bias of BPNN.Then use the Genetic Neural Network to predict the wind speed,the result shows that GA-BP is a more effective and accurate than Time Series.Finally,this paper makes use of the relationship between wind speed and power generation to predict the wind turbine power generation.