Structural Health-Monitoring and Assessment in Tunnels: Hybrid Simulation Approach

Abstract

During the operation of tunnels, structural performance could inevitably degrade due to stochastic and adverse factors. To reduce the randomness and uncertainty of the tunnel operation, this paper provides a novel global sensitivity analysis (GSA) approach for monitoring parameters. The Wuhan Yangtze River Tunnel was utilized as a case study to verify the applicability of the proposed approach. The particle swarm optimization–least-squares support vector machine (PSO-LSSVM) was used to establish the model relationship between the input and output parameters, and the variance-based extended Fourier amplitude sensitivity test (EFAST) algorithm was employed to investigate the parameters sensitivities. Results of this GSA quantified the parameter sensitivities, and the sensitive and insensitive parameters were distinguished. The sensitive parameters can be identified as major factors for structural health monitoring and proactive maintenance in tunnels. This study evaluated the variations of sensitivity index with various target functions, parameter ranges, and distributions. The variations of parameter sensitivities were observed under various conditions, which indicated that sensitive and insensitive parameters may be different under different conditions. The GSA in this research can aid in optimizing tunnel design, construction, and operating period safety management.