针对风功率难以预测的问题,提出一种基于小波分解和ELMAN神经网络的风速-风功率预测模型,采用小波分解来降低风速的非平稳性;采用ELMAN神经网络建立风速预测模型;基于实测数据拟合功率曲线,并结合得到的功率曲线进行风功率预测.最后将建模流程应用到实测数据验证模型的有效性,结果表明了模型的先进性.
In the light of the difficulty of wind power forecasting, the paper proposes a kind of prediction model of wind speed-power based on wavelet-ELMAN neural network, First, wavelet decomposition is adopted to reduce the non-stationary of wind speed; Then the prediction model of wind speed is built based on ELMAN neural network; Finally, the wind power prediction is carried out combining with the power curve from the fitting curve of measured data, and the advanced nature of the proposed model is validated by applying this modeling process to the measured data.