Scientific Reports

Table 6 Displaying the time, in seconds, required to converge to a total loss value of 10^− 6.

From: Enhancing convergence speed with feature enforcing physics-informed neural networks using boundary conditions as prior knowledge

	Ratio	FE-PINN	Vanilla PINN
	Ratio	FE-PINN	\(\:\lambda\:=1\)	\(\:\lambda\:=1.2\)	\(\:\lambda\:=1.4\)	\(\:\lambda\:=1.6\)	\(\:\lambda\:=1.8\)
Case 1	6.25	6 + 46	81	86	92	95	110
Case 2	6.25	4 + 32	68	65	72	72	80
Case 3	6.25	4 + 42	71	82	82	83	88
Case 1	8.75	4 + 46	72	64	67	74	79
Case 2	8.75	6 + 61	72	78	88	82	73
Case 3	8.75	6 + 35	47	55	57	56	62
Case 1	11.25	5 + 34	59	57	60	61	64
Case 2	11.25	6 + 40	79	89	92	92	91
Case 3	11.25	5 + 61	67	83	96	95	96
Case 1	15	5 + 49	93	98	106	121	123
Case 2	15	5 + 60	72	89	92	97	101
Case 3	15	5 + 28	81	89	96	109	119
Case 1	18.75	5 + 82	91	96	101	106	114
Case 2	18.75	5 + 39	105	107	112	139	135
Case 3	18.75	4 + 41	106	106	122	130	134

Back to article page

Search

Advanced search

Quick links