超短期风电功率预测的混合深度学习模型

针对风电功率预测(WPF)问题,提出一种基于离散小波变换(DWT)、时间卷积网络(TCN)和长短期记忆(LSTM)神经网络的混合深度学习模型(DWT-TCN-LSTM),对超短期风电功率进行预测.将DWT-TCN-LSTM模型分别与差分整合移动平均自回归(ARIMA)模型,支持向量回归(SVR)模型,长短期记忆神经网络模型和卷积长短期记忆(TCN-LSTM)混合模型进行对比实验,通过对称平均绝对百分比误差(SMAPE),均方根误差(RMSE)和平均绝对误差(MAE)3种评价指标值对各个模型进行评价.实验结果表明:DWT-TCN-LSTM模型具有较好的预测性能.

Hybrid Deep Learning Model Based on Super-Short-Term Wind Power Forecasting

Aiming at the problem of wind power forecasting(WPF), a hybrid deep learning model(DWT-TCN-LSTM)based on discrete wavelet transform(DWT), time convolutional network(TCN)and long and short-term memory(LSTM)neural network is proposed to predict the super-short-term wind power. The DWT-TCN-LSTM model is compared experimentally with the differential integrated moving average autoregressive model(ARIMA), support vector regression(SVR)model, long and short-term memory neural network model and convolutional long and short-term memory(TCN-LSTM)mixed model. The each model is evaluated through three evaluation metrics of symmetric mean absolute percent error(SMAPE), root mean square error(RMSE)and mean absolute error(MAE). The experimental results show that: the DWT-TCN-LSTM model has better prediction performance.