DIAGNOSIS OF CONCURRENT SENSOR FAULTS IN STRUCTURAL HEALTH MONITORING SYSTEMS

Abstract

Sensor faults in structural health monitoring (SHM) systems may occur due to aging, exposure to harsh weather conditions, manufacturing defects in hardware components, damage during installation or operation, and issues with data transmission. If undetected, sensors faults may result in inaccurate or incomplete sensor readings, which may significantly impact the accuracy, reliability, and performance of SHM systems. As a result, fault diagnosis in SHM systems may help improve the accuracy, reliability, and performance of SHM systems. However, most fault diagnosis approaches for SHM only consider single-fault occurrence, which may oversimplify actual fault occurrences in real-world SHM systems, where sensor faults may occur concurrently in multiple sensors. To extend fault diagnosis in SHM towards concurrent sensor faults in multiple sensors, this paper presents an adaptive fault diagnosis approach based on analytical redundancy. The approach encompasses four steps, (i) initialization (ii) fault detection, (iii) fault isolation and (iv) fault accommodation, using correlated data from multiple sensors of an SHM system. The proposed fault diagnosis approach is validated using data recorded using a real-world SHM system. The results show the high accuracy, reliability, and performance of the proposed approach in detecting concurrent sensor faults in real-world SHM systems.