Gross outlier removal and fault data recovery for SHM data of dynamic responses by an annihilating filter‐based Hankel‐structured robust PCA method

Abstract

In daily monitoring of structures instrumented with long-term structural health monitoring (SHM) systems, the acquired data is often corrupted with gross outliers due to hardware imperfection and/or electromagnetic interference. These unexpected spikes in data are not unusual and their existence may greatly influence the results of structural health evaluation and lead to false alarms. Hence, there is a high demand for executing data cleaning and data recovery, especially in harsh monitoring environment. In this paper, we propose a robust gross outlier removal method, termed Hankel-structured robust principal component analysis (HRPCA), to remove gross outliers in the monitoring data of structural dynamic responses. Different from the deep-learning-