基于神经网络平均影响值的超短期风电功率预测

Ultra-short-term Wind Power Prediction Based on Neural Network and Mean Impact Value

针对动态神经网络风电功率预测模型输入变量较多、模型复杂的问题,将神经网络和平均影响值方法相结合,提出了一种基于神经网络平均影响值的超短期风电功率预测方法。此方法综合考虑了各输入变量对输出变量(风电预测功率)的外部贡献率和内部贡献率,筛选出了对输出变量贡献率最大的输入变量,建立了一个优化的神经网络超短期风电功率预测模型。实验结果表明,所提模型降低了预测模型的复杂度,减少了测量噪声对预测精度的影响,得到了较好的风电功率预测结果。

To solve the problems of variable redundancy and model complexity in the prediction model based on the dynamic neural network,an ultra-short-term wind power prediction method is proposed by combining the neural network（NN）and the mean impact value（MIV）.In this method,the external and internal contribution rates of the input variables to the output variables（wind power prediction value）are taken into account,and the input variable with the largest contribution to the output variables is selected.Then an optimized NN prediction model for ultra-short-term wind power prediction is developed.The experimental results show that the proposed model reduces the complexity of the prediction model,mitigates the influence of the measuring noise on the prediction accuracy,and obtains good wind power prediction results.