Open Access

表4

伪动态工况不同训练集多步预测量化结果

训练集
占比/%
预测方法 ERMS EMA R2
50 LSTM 58.4 35.9 0.679
Transformer 1.11 0.728 0.999
LSTM-TS 7.98 2.85 0.994
CNN-LSTM 25.5 15.7 0.939
CNN 14.7 8.29 0.98
60 LSTM 39.9 23.9 0.863
Transformer 1.24 0.834 0.999
LSTM-TS 9.4 2.74 0.992
CNN-LSTM 15.1 12.6 0.98
CNN 22.3 10 0.957
70 LSTM 32.9 18.7 0.877
Transformer 1.85 1.31 0.999
LSTM-TS 1.89 1.13 0.999
CNN-LSTM 5.03 3.27 0.997
CNN 9.11 7.05 0.991
80 LSTM 33 19.2 0.798
Transformer 2.07 1.32 0.999
LSTM-TS 2.01 1.34 0.999
CNN-LSTM 6.8 5.09 0.991
CNN 7.83 5.27 0.989

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.