A Novel Prediction Model for Car Body Vibration Acceleration Based on Correlation Analysis and Neural Networks

Table 6

Performances of the FFNN-, TSNF-, and TSF-structure models on the new dataset.

Method	FF			TSNF			TSF
	MSE	R	MAE	MSE	R	MAE	MSE	R	MAE
BQ	1.31e-03	-0.37	3.07	6.61e-04	0.2305	0.0213	4.83e-04	0.61	1.86
CGB	1.30e-03	-0.16	2.97	0.0020	0.3609	0.0413	1.06e-04	0.78	0.81
CGF	7.14e-03	-0.25	7.95	0.0029	0.4036	0.0376	9.35e-04	0.95	2.76
CGP	1.01e-03	0.39	2.84	0.0011	0.2284	0.0275	1.53e-04	0.91	1.03
OSS	4.28e-03	0.09	5.63	0.0011	-0.0584	0.0275	1.71e-03	0.80	3.79
RB	2.35e-03	-0.06	3.61	0.0021	0.2364	0.0351	4.50e-04	0.42	1.72
SCG	3.85e-03	0.69	4.36	3.05e-04	0.0131	0.0128	4.06e-04	0.79	1.64
LM	1.02e-03	0.66	2.76	0.0122	-0.3984	0.0864	5.50e-03	0.04	4.52
BR	3.74e-04	0.68	1.27	0.0111	-0.0890	0.0696	1.01e-03	0.84	2.93