Bayesian Probabilistic Framework for Damage Identification of Steel Truss Bridges under Joint Uncertainties

Table 4

Identified damages based on model class M3 with three measurement schemes.

Number of sensor measurements^a	Damage index 𝜽𝑖 of different truss members and its COV^b				Standard deviation of prediction error 𝜎^2 (COV)
	𝛼1 (COV); Member G74	𝛼2 (COV); Member G75	𝛼3 (COV); Member G83	𝛼4 (COV); Member G84
4-sensor; (nodes 5–8)	0.8880; (11.189%)	0.9202; (16.211%)	0.8612; (11.318%)	0.9075; (9.6544%)	47.861; (1.0072%)
8-sensor; (nodes 5–8, 15–18)	0.8938; (10.827%)	0.8263; (10.120%)	0.8884; (9.785%)	0.9250; (8.692%)	56.838; (1.016%)
10-sensor; (5–8, 15–18, 22, and 31)	0.8370; (7.1801%)	0.8527; (7.3334%)	0.8602; (8.3060%)	0.9507; (5.8878%)	37.886; (0.4888%)
True value	0.85	0.85	0.85	1.00	—

Note: a: the measure point is referred to as sensor deployment schemes in Table 2.
b: COV is the coefficient of variation of identified damage index and is listed in parenthesis below.