Vibration monitoring of civil engineering structures has gained a lot interest over the past few years due to relative ease of instrumentation and the development of powerful system identification techniques. The main idea is to replace visual, systematic inspections by health monitoring systems which continuously acquire and analyze vibration data and allow to identify damage in an early stage. This paper discusses the application of an approach to determine that damage is present as well as the geometric location. Hereto, a statistical local approach based on covariance-driven stochastic subspace identification is proposed. The approach is applied to vibration data measured on the bridge Z24 in Switzerland. Ambient responses data were measured right before and after applying a damage pattern to the Z24 bridge.
展开▼
