Identification of influencing factors on bridge damages using Bayesian network

Abstract In Japan, bridge inspections are compulsorily performed in 5‐year cycles. With the institutionalization of inspection cycle, essential data have been continuously accumulated. However, effective utilization requires trend analysis and causal for a group bridges. this study, method determining factors affecting deterioration is established. The concrete steel bridges with Bayesian networks by utilizing on repair, open such as traffic census rainfall. For bridges, target members deck slab main structural members, whereas damage type “Delamination/rebar exposure” “corrosion,” respectively. validity selected explanatory variables verified crossvalidation using separately prepared test data; evidently, maximum rating prediction accuracy 86%. Furthermore, influencing extracted study reasonable two damages, thus indicating possibility probabilistically extracting specific damages networks.