Mining high-frenquncy data and its application to structural health monitoring

Structural health monitoring deploys various types of sensors on a structure to monitor the health status. The sensor data are high-frequency heterogeneous data, so a massive amount of data are generated each day. Our research aims to detect anomalies and to evaluate the health status of a structure online. This PhD project proposes four approaches to handle online anomaly detection and structural health evaluation, and these methods have been verified through empirical evaluations with public datasets and practical datasets. The proposed approaches help civil engineering field to identify risky circumstances early and to develop maintenance plans and recovery plans efficiently