Studies on data correction of Structural Health Monitoring using ultrasonic guided waves: case of study
Résumé
Structural health monitoring (SHM) aims to rule on the integrity of a given structure. This can be reached by means of various nondestructive testing techniques. SHM is based on sensors and actuators that are designed to live permanently with the structures to be monitored. These transducers should be interrogated regularly during a period long enough to build a database. These collected data are unfortunately vulnerable to environmental and operational conditions changes (EOC such as temperature, humidity, pressure, vibration, flow, etc.), which can influence the used electronic devices, transducers time behavior, and even the structure itself. In order to get a reliable health monitoring, data should be corrected (i. e. minimizing the error induced by these EOC). In general, data correction can be manual or automatic, but in SHM systems, it should be automatic and so, algorithms should be developed to resolve efficiently this problem. The current communication deals with this task. It presents an extensive description of the aforementioned problem, and literature review of the methods developed right now in literature. It describes the technique of data correction proposed by the authors. This technique has been applied on a case of study which was a pipeline. Main results will be shown and debated.