Wind Power Forecasting And The Maximum Wind Energy Tracking Of Wind Power Unit

With the reduction of stored energy and increasing environmental pollution, the adoption of non-polluting energy has become an important approach to solve these problems. Wind energy which is clean and renewable is providing more energy in the proportion of total electricity power. However, due to the intermittent nature of wind power and uncontrollable of other characteristics, the ability to accept wind power has met the bottleneck of the development on wind power. More accurate predictions of wind power which will improve the electricity market competitiveness of wind field can effectively reduce and avoid the harmful effects of wind field on power system. Meanwhile, the development of more capacity and’ more effective wind turbine generator is needed to improve the utilization of wind energy which makes the study of the maximum wind energy tracking becoming significantly important.In this paper, the practical problems in wind power generation concerning wind power prediction and maximum wind energy tracking has been studied deeply. This paper provided a forecasting approach named as Wavelet Time-seires BP neural network approach containing wavelet theory, time series theory and neural network theory. Firstly, by wavelet decomposition, the original wind speed waveform is decomposed into two parts, one is a trend signal of low frequency and the other contains five random signals of high frequency. Then, the trend signal is predicted via BP neural network. And the random signals are predicted through training BP neural networks respectively, in which the input variables are picked out by the time series method. Thirdly, the predictions of the six signals above are added to obtain the prediction result of the original wind speed signal. At last, the predicted values are used as input of the wind turbine power curve to forecast the wind power. The measured data of a wind power plant in south China is used to test the accuracy of the proposed approach in simulation. In other part of this paper, controlling method was obtained to achieve the maximum wind power tracking by analyzing the characteristics of wind turbine. To achieve the active reactive power decoupling control of generator, the vector transformation technology of stator-flux oriented was adopted and characteristic of doubly fed induction generator was considered. Consequently, the purpose to reach the maximum wind power tracking was achieved. The results proved the correction of model by applying Matlab/Simulink which conducted the complete wind power generation system simulation.