Impact of Data Partitioning to Improve Prediction Accuracy for Remaining Useful Life of Li-Ion Batteries

Table 2

Average prediction accuracy for each of data partitioning methods.

Method	Score	LFP	NCA	NMC	NCA and NMC	ALL
RND	MAE (cycle)	3.2	1.4	3.2	2.8	3.1
	RMSE (cycle)	18.8	4.5	13.1	12.6	17.5
	MAPE (%)	0.0049	0.0250	0.0276	0.0288	0.0166
SEP	MAE (cycle)	111	35	139	160	173
	RMSE (cycle)	134	40	155	234	223
	MAPE (%)	0.357	0.823	1.590	1.112	0.453
INI	MAE (cycle)	56	47	38	—	—
	RMSE (cycle)	66	53	43	—	—
	MAPE (%)	0.557	2.002	1.855	—	—
BOX	MAE (cycle)	75	55	49	—	—
	RMSE (cycle)	86	63	56	—	—
	MAPE (%)	0.463	2.527	2.496	—	—
ADD	MAE (cycle)	54	47	39	—	—
	RMSE (cycle)	64	53	44	—	—
	MAPE (%)	0.671	2.026	1.890	—	—