Hybrid aI-based sensor optimization for structural health monitoring of multi-story buildings — case study : HQ tower R+12

Abstract

Structural Health Monitoring (SHM) plays a critical role in ensuring the safety and func- tionality of vital structures such as bridges, dams, and public buildings. To make SHM sys-tems more cost-effective, it is essential to optimize the number and placement of sensors, reducing implementation costs while maintaining reliable damage detection and structural assessment. This study addresses the problem of optimizing damage-adaptive sensor layout in struc- tural health monitoring (SHM) for multi-story buildings. A three-dimensional finite ele- ment model of a 12-story reinforced concrete tower (R+12) was developed using SAP2000, enabling the identification of high-risk damage zones based on internal force distributions. Steady-state vibration responses were generated, and a genetic algorithm was used to iden-tify the optimal sensor configuration for each damage scenario using distance matrices as damage-sensitive features. These scenario-based layouts were then merged into a unified configuration by analyzing sensor occurrence and importance scores. The final sensor set ensures sufficient coverage and sensitivity to structural degradation while maintaining a reduced number of sensors. The proposed approach provides a scalable and practical solution for SHM system design in complex structures with anticipated damage regions