基于mRMR和VMD-AM-LSTM的短期风功率预测

Short-term Wind Power Prediction Based on mRMR and VMD-AM-LSTM

为了提高模型预测风功率的准确率,提出了一种基于最大相关-最小冗余筛选、变分模态分解、注意力机制和长短期记忆神经网络的短期风功率预测方法。首先使用变分模态分解算法将风功率序列分解成几个中心频率不同的分量;再对各个分量结合最大相关-最小冗余筛选出的气象特征分别建立注意力机制和长短期记忆混合预测模型;最后将各个分量的预测结果叠加,得到最终的风功率。实际算例表明,与其他几种模型对比,所提预测方法准确率明显提升。

In order to improve the accuracy of the model to predict wind power,a short-term wind power prediction method based on maximum correlation-minimum redundancy filtering,variational mode decomposition,attention mechanism and long and short-term memory(LSTM) neural network is proposed.First,the variational mode decomposition algorithm is used to decompose the wind power sequence into several components with different center frequencies.Then,an attention mechanism and a mixed prediction model of long and short-term memory are established for each component combined with the meteorological characteristics selected by the maximum correlation-minimum redundancy.Finally,the prediction results of each component are superimposed to obtain the final wind power.Through practical examples,it is shown that compared with several other models,the prediction accuracy of the proposed method is significantly improved.