基于重组二次分解及LSTNet-Atten的短期负荷预测

刘洪伟,王磊,刘阳,张鹏超,乔石

Short term load forecasting based on recombination quadratic decomposition and LSTNet-Atten

Hongwei LIU,Lei WANG,Yang LIU,Pengchao ZHANG,Shi QIAO

表 8 提出模型与混合模型的预测结果对比

Tab.8 Comparison of prediction result of proposed model and hybrid model

预测方法 R2 MAPE/% MAE/MW RMSE/MW 本文模型 0.9941 0.9352 288.2614 344.4416 LSTGru-Atten 0.9923 1.0706 329.5825 389.6542 LSTNet-Attention 0.9910 1.1043 343.4469 410.2345 CNN-LSTM 0.9888 1.3307 409.6971 472.1109 CNN-GRU 0.9892 1.2605 385.6182 465.2339 TCN-LSTM 0.9892 1.2345 379.9722 462.5232 TCN-GRU 0.9900 1.1714 362.9603 447.8847