Missing measurement data recovery methods in structural health monitoring: The state, challenges and case study

Abstract

In the field of structural health monitoring (SHM), the sensor measurement signals collected from the structure are the foundation and key of the SHM system. However, the loss of sensor measurement signals can affect the accurate assessment of structural health. The restoration of missing measurement signals in SHM is a multidisciplinary research field. Therefore, analyzing the features of the measurement signals from multiple perspectives, establishing appropriate mathematical models, and selecting efficient algorithms is crucial to solving this problem. This article briefly reviews the latest research progress on restoring missing sensor measurement signals in SHM, using mathematical models as classification criteria, including finite element methods, sparse representation methods, statistical inference methods, and machine learning algorithms. At the end of this article, a study is conducted on an engineering case, and the development trend and challenges of restoring missing measurement sensor signals in SHM are presented from multiple perspectives in-depth.