基于改进LSTM的商业建筑冷负荷预测模型

董芳楠,武强,刘佳瑶,于军琪

Cooling load prediction model for commercial buildings based on improved LSTM

Fangnan DONG,Qiang WU,Jiayao LIU,Junqi YU

表 6 不同模型对不同月份建筑物冷负荷预测准确性的对比

Tab.6 Comparison of monthly building cooling load prediction accuracy across different models

建筑 月份 模型 CV-RMSE 建筑 月份 模型 CV-RMSE 建筑1 6月 LSTM 1.387 建筑2 6月 LSTM 1.197 SVR 2.659 SVR 2.114 BPNN 3.019 BPNN 2.546 Adam-LSTM 1.113 Adam-LSTM 1.000 SCOA-LSTM 0.641 SCOA-LSTM 0.641 WAdam-LSTM 0.761 WAdam-LSTM 0.833 7月 LSTM 1.812 7月 LSTM 1.806 SVR 2.422 SVR 2.359 BPNN 2.767 BPNN 2.871 Adam-LSTM 1.066 Adam-LSTM 1.057 SCOA-LSTM 0.867 SCOA-LSTM 0.812 WAdam-LSTM 0.653 WAdam-LSTM 0.487 8月 LSTM 1.803 8月 LSTM 1.707 SVR 2.387 SVR 2.308 BPNN 2.281 BPNN 2.436 Adam-LSTM 1.141 Adam-LSTM 1.108 SCOA-LSTM 1.478 SCOA-LSTM 1.337 WAdam-LSTM 0.592 WAdam-LSTM 0.595