Table 5 Comparison of SOH prediction performance.

From: Estimation of lithium-ion battery health state using MHATTCN network with multi-health indicators inputs

Battery ID	Training range	Model	Evaluation metrics
Battery ID	Training range	Model	MAE	MSE	\(R^{2}\)	RMSE	MAPE
CS2_35	200	MHAT-TCN(m)	0.0124	0.0011	0.9750	0.0329	2.4682
		MHAT-TCN	0.0229	0.0011	0.9743	0.0333	3.1102
		TCN	0.0111	0.0018	0.9581	0.0425	2.3806
	400	MHAT-TCN(m)	0.0129	0.0006	0.9868	0.0239	2.2358
		MHAT-TCN	0.0124	0.0007	0.9827	0.0273	2.4656
		TCN	0.0271	0.0012	0.9729	0.0342	3.7029
	600	MHAT-TCN(m)	0.0106	0.0003	0.9926	0.0178	1.7622
		MHAT-TCN	0.0120	0.0004	0.9899	0.0208	2.1789
		TCN	0.0242	0.0010	0.9769	0.0316	3.3214
	800	MHAT-TCN(m)	0.0054	0.0001	0.9973	0.0107	0.9128
		MHAT-TCN	0.0089	0.0003	0.9940	0.0161	1.7012
		TCN	0.0155	0.0008	0.9819	0.0279	2.4160
CS2_36	200	MHAT-TCN(m)	0.0142	0.0006	0.9898	0.0262	3.6990
		MHAT-TCN	0.0350	0.0015	0.9779	0.0385	4.6902
		TCN	0.0425	0.0021	0.9690	0.0455	6.0991
	400	MHAT-TCN(m)	0.0199	0.0005	0.9916	0.0237	3.2740
		MHAT-TCN	0.0283	0.0010	0.9849	0.0318	4.7726
		\({\textrm{TCN}}\)	0.0416	0.0020	0.9709	0.0442	6.0481
	600	MHAT-TCN(m)	0.0153	0.0003	0.9941	0.0199	2.7109
		MHAT-TCN	0.0221	0.0007	0.9899	0.0259	3.5215
		TCN	0.0349	0.0014	0.9786	0.0379	5.4394
	800	MHAT-TCN(m)	0.0100	0.0002	0.9958	0.0167	2.1303
		MHAT-TCN	0.0192	0.0005	0.9924	0.0225	2.8885
		\({\textrm{TCN}}\)	0.0285	0.0011	0.9845	0.0323	4.5282
CS2_37	200	MHAT-TCN(m)	0.0111	0.0006	0.9862	0.0243	2.3264
		MHAT-TCN	0.0116	0.0009	0.9792	0.0299	2.5251
		TCN	0.0233	0.0014	0.9664	0.0379	4.8994
	400	MHAT-TCN(m)	0.0119	0.0004	0.9892	0.0215	2.1830
		MHAT-TCN	0.0133	0.0006	0.9857	0.0249	2.4447
		TCN	0.0200	0.0008	0.9818	0.0280	3.3429
	600	MHAT-TCN(m)	0.0104	0.0003	0.9924	0.0181	1.8947
		MHAT-TCN	0.0129	0.0005	0.9896	0.0212	2.4677
		TCN	0.0128	0.0007	0.9829	0.0271	2.6678
CS2_38	200	MHAT-TCN(m)	0.0212	0.0006	0.9837	0.0260	3.1114
		MHAT-TCN	0.0362	0.0057	0.8606	0.0761	7.1129
		\({\textrm{TCN}}\)	0.0417	0.0067	0.8389	0.0818	7.9465
	400	MHAT-TCN(m)	0.0135	0.0004	0.9907	0.0196	2.1761
		MHAT-TCN	0.0219	0.0009	0.9802	0.0287	3.4360
		\({\textrm{TCN}}\)	0.0240	0.0009	0.9782	0.0442	3.8260
	600	MHAT-TCN(m)	0.0105	0.0003	0.9937	0.0162	1.5603
		MHAT-TCN	0.0118	0.0005	0.9873	0.0230	2.8160
		\({\textrm{TCN}}\)	0.0221	0.0009	0.9799	0.0289	3.5162
	800	MHAT-TCN(m)	0.0062	0.0002	0.9962	0.0125	1.0319
		MHAT-TCN	0.0107	0.0003	0.9928	0.0173	1.6150
		\({\textrm{TCN}}\)	0.0195	0.0006	0.9846	0.0253	2.9505

Back to article page

Table 5 Comparison of SOH prediction performance.

Search

Quick links