Robustness-oriented optimal sensor placement for structural monitoring considering sensor failures

Abstract

Conventional optimal sensor placement (OSP) methods employ the premise that all sensors work perfectly during long-term structural monitoring. However, this premise is often difficult to fulfill in real applications due to poor manufacturing and material aging of sensors, human damage, and electromagnetic interference. This paper presents a robustness-oriented OSP method that considers sensor failures. The OSP problem is designed with consideration of sensor failures to ensure that both complete vibration data collected by all sensors and incomplete vibration data caused by individual sensor failures can accurately identify structural modal parameters. A dispersion-aggregation firefly algorithm (DAFA), which is derived from the basic firefly algorithm, has been proposed to solve this complicated optimization problem. The dispersion and aggregation operators are designed to prevent falling into local optima and to rapidly converge to the global optima. The proposed methodology is confirmed by extracting the robust sensor configuration for a long-span cable-stayed bridge. The robustness of the optimal sensor configurations against sensor failure is thoroughly explored, and the performance of the proposed DAFA is extensively examined.

Keywords

structural health monitoring (SHM) / optimal sensor placement (OSP) / long-span bridges / modal parameter identification / firefly algorithm

Funding

National Natural Science Foundation of China(51978243)

National Natural Science Foundation of China(52578360)