A Prediction Network for Hydraulic Support Pressure Based on Multitimescale Feature Extraction

Table 3

Errors in the prediction results of each model.

Model	Hydraulic support no. 25		Hydraulic support no. 45		Hydraulic support no. 65		Hydraulic support no. 85
	RMSE (kN)	MAE (kN)	RMSE (kN)	MAE (kN)	RMSE (kN)	MAE (kN)	RMSE (kN)	MAE (kN)
Ours	408.63	255.24	509.74	371.21	576.10	422.14	564.88	381.38
GRU	416.98	259.94	600.49	411.05	602.35	436.61	582.89	390.02
ANN-GA	523.89	299.35	616.99	432.32	720.14	504.18	767.98	503.39
TCN	592.73	360.21	643.98	499.22	812.22	571.66	663.33	428.58
DNN-BTF	436.41	272.93	546.67	373.81	659.62	458.23	608.07	395.71
LSTM
+SVM	1994.68	571.24	2594.39	690.22	2749.47	696.74	2587.59	663.45

The bold values are intended to mark the values with the smallest errors in the comparison experiments.