Diagnosis of simultaneous sensor faults in structural health monitoring systems

Abstract: Current fault diagnosis concepts in structural health monitoring (SHM) are limited by the sole consideration of single sensor faults, which does not fully capture the complexity of real-world faults in SHM systems. This paper presents an adaptive fault diagnosis approach for SHM systems that addresses multiple sensor faults simultaneously. The approach, based on analytical redundancy, includes fault detection, isolation, and accommodation, and has been validated using real-world sensor data recorded from a railway bridge. The results show the high accuracy, reliability, and performance of the proposed approach regarding multiple sensor faults that occur simultaneously in real-world SHM systems.