Aero Engine Gas-Path Fault Diagnose Based on Multimodal Deep Neural Networks

Table 2

Comparison of experimental results on UCR standard datasets.

Dataset	Proposed	NN-based methods					Distance-based methods
		ResNet	MCNN	CTN-T	ESN	TN	ST	LWDTW	BOSS
Chlor.Conc	0.999	0.84	0.797	0.83	0.920	0.731	0.700	0.644	0.66
Cinc_ECG	1.00	—	0.942	—	0.679	—	0.846	0.935	0.901
Dist.phal.O.C	0.864	0.80	—	0.80	—	0.812	—	0.739	0.815
ECGFivedays	1.00	0.97	1.00	1.00	1.00	0.926	0.999	0.835	0.983
yoga	0.941	0.87	0.888	0.92	0.820	0.84	0.846	0.847	0.901
average	0.961	0.88	0.907	0.91	0.855	0.827	0.838	0.800	0.852