Damage Diagnosis Of Large-span Bridges Based On Multi-grade Modal Parameter Identification

It is a great challenge to diagnose the condition of long-span bridges by using the massive monitoring data owing to the difficulty of eliminating the effect of the environmental factors and other loads on massive monitoring data. To overcome this issue, this study focused on damage diagnosis of long-span bridges by using the massive monitoring data.Recent research results and development trend of damage detection of bridges considering the effect of environmental factors were summarized systematically, and some key issues of damage detection of bridges using massive monitoring data were analyzed. Based on these works, the idea of classification of modal parameters was proposed according to different effective extent of environmental factors on modal parameters.Multi-level modal parameter identification was paid much attention in this study. The environmental factors, such as traffic, air temperature, air humidity and wind etc., could be classified into one class by following a certain rule. Under this class, the effect of different environmental factors on modal parameters was similar, so the statistical rule or characteristics of modal parameters belonging to this class could be obtained by statistical analysis. Actually, this method was a practical way to deal with the effect of environmental factors on modal parameters of long span bridges.Based on multi-level modal parameter identification, the relationship between modal parameters and environmental factors was determined under each class. With the whole frame of novelty detection, a damage detection method based on multi-level modal parameter identification was proposed.With a practical bridge, No. 3 Nanjing Yangzi River Bridge, the effectiveness of proposed method was demonstrated by analyzing the changing rule of massive monitoring modal parameters acted on massive monitoring environmental data.