Vibration-based BHMS for long-span bridges considering environmental actions

BHMSs(Bridge Health Monitoring Systems) have been widely developed and applied to actual bridges in recent years because monitoring and maintenance of bridges have become more critical issues. The final target of applying a BHMS may be to monitor the health condition of a bridge on a regular basis and to detect any abnormal condition change in advance to possible hazardous accidents. Major components of a BHMS may include (a) data acquisition, (b) data processing, and (c) evaluation of the bridge health. In addition, depending on the owner's request, intervention should be suggested based on the evaluated health condition. However, since bridges are highly influenced by environmental actions such as temperature variation, it is difficult to extract pure bridge response from the measured data. Especially long-span bridges are suspected to be more seriously influenced by environmental actions. In the paper, three steps for BHMS are proposed: (a) application of some numerical schemes to recover or correct loss data, (b) application of some statistical methods to eliminate environmental effects or outliers from the data, and (c) computation of the health index to evaluate the health condition of the bridge. The proposed BHMS algorithm is examined by using long-term measured data from a long-span cable-stayed bridge.