A Fault Identification Method for Electric Submersible Pumps Based on DAE-SVM

Table 8

.Comparison of different methods in evaluation index.

Methods		Conditions
		Condition 1	Condition 2	Condition 3	Condition 4	Average value
DAE-SVM	MAE	0.0289	0.0306	0.0457	0.0363	0.0354
	RMSE	0.0320	0.0543	0.0461	0.0231	0.0389
	MAPE	0.1559	0.2474	0.1339	0.3579	0.2238
	R^2 (%)	95.33	96.71	94.33	97.34	95.93
SVM	MAE	0.0268	0.0359	0.0475	0.0512	0.0403
	RMSE	0.0421	0.0435	0.0641	0.0631	0.0532
	MAPE	0.2559	0.5474	0.3339	0.2579	0.3488
	R^2 (%)	96.33	93.71	91.33	92.34	93.43
KNN	MAE	0.0368	0.0451	0.0395	0.0212	0.0357
	RMSE	0.0368	0.0569	0.0577	0.0571	0.0521
	MAPE	0.0168	0.0259	0.0475	0.0562	0.0366
	R^2 (%)	94.33	92.71	92.33	96.34	93.93
RF	MAE	0.0489	0.0572	0.0699	0.0469	0.0557
	RMSE	0.0268	0.0469	0.0498	0.0757	0.0498
	MAPE	0.0168	0.0259	0.0475	0.0562	0.0366
	R^2 (%)	91.33	94.73	95.12	93.72	93.73
XGB	MAE	0.0227	0.0401	0.0427	0.0437	0.0373
	RMSE	0.0414	0.0661	0.0397	0.0358	0.0458
	MAPE	0.2510	0.4221	0.3137	0.4271	0.3535
	R^2 (%)	94.62	94.82	95.12	94.55	94.78