Damage Identification of a Steel Frame Based on Integration of Time Series and Neural Network under Varying Temperatures

Table 9

The identification results of damage severities when one brace is damaged with noises.

Damaged brace	aj (%)	T (°C)	Target output	Noise intensity (1%)		Noise intensity (3%)
				Actual output	Relative error (%)	Actual output	Relative error (%)
1	50	20	0.5	0.457	8.57	0.48937	−2.13
		30	0.5	0.534	−6.74	0.52943	5.89
	70	20	0.7	0.712	−1.76	0.6847	−2.19
		30	0.7	0.727	−3.91	0.7735	10.50
4	50	20	0.5	0.492	1.50	0.47049	−5.90
		30	0.5	0.468	6.39	0.45675	−8.65
	70	20	0.7	0.769	−9.81	0.76837	9.77
		30	0.7	0.743	−6.15	0.71513	2.16
5	50	20	0.5	0.491	1.74	0.41528	−16.94
		30	0.5	0.528	−5.53	0.41918	−16.16
	70	20	0.7	0.740	−5.67	0.72871	4.10
		30	0.7	0.738	−5.40	0.72388	3.41