基于选择性深度神经网络集成的涡扇发动机剩余寿命预测

韩冬阳,林泽宇,郑宇,郑美妹,夏唐斌

Remaining useful life estimation of turbofan engine based on selective ensemble of deep neural networks

Dong-yang HAN,Ze-yu LIN,Yu ZHENG,Mei-mei ZHENG,Tang-bin XIA

表 6 不同预测方法精度对比

Tab.6 Comparison results with methods proposed

预测方法 RMSE SCORE FD001 FD003 FD001 FD003 Deep-CNN[6] 12.61 12.64 274.00 284.10 Deep-LSTM[25] 16.14 16.18 338.00 852.00 CNN-LSTM[26] 16.13 17.12 303.00 1 420.00 RNN-SPI[27] 13.58 19.16 228.00 1 727.00 MODBNE[15] 15.04 19.41 334.00 683.40 RULCLIPPER[28] 13.27 16.00 216.00 317.00 选择性神经网络集成 12.00 13.08 282.00 314.80