correlated damage probabilities of bridges using multinomial logistics regression model in seismic risk assessment of transportation networks

In this paper, using the multinomial models, the spatial correlation of the damage probabilities among different nodes in lifeline networks is assessed. in addition to the correlation of the groundmotion intensities, the cross-correlation of the damage probabilities in nodes of a lifeline network and the cross-correlation of five damage states (i.e. non-damage, slight, moderate, extensive and complete) in each node are considered. in this paper, a methodology is developed to calculate the damage probabilities of a bridge by considering the damage probabilities of other bridges. moreover, the effect of the damage probability on the edtt (extra daily traffic time) in a transportation network is investigated as an operational parameter.