Document Type : Regular Article
Authors
1
civil , engineering, Kharazmi University
2
School of Engineering Kharazmi
3
Civil Engineering Department, Engineering Faculty, Kharazmi University, Tehran, Iran
4
Faculty of Civil Engineering, Kharazmi University
10.22115/scce.2024.429581.1763
Abstract
Super-tall skyscrapers are crucial for modern urbanization, but ensuring their safety in natural events is vital. Structural Health Monitoring (SHM) is essential for assessing the condition of these iconic structures. This study focuses on the Milad Tower in Tehran, Iran, and the Canton Tower in Guangzhou, China, aiming to detect, locate, and assess damage severity. The goal is to detect damage, locate affected areas, and assess severity. The following techniques are employed: a) Support Vector Machines (SVM), b) Multivariate Empirical Mode Decomposition (MEMD), c) Wavelet Transform (WT). SVMs efficiently identify damage patterns. However, require parameter tuning, addressed using Observer-Teacher-Learner-Based Optimization. MEMD interprets signals well, allowing simultaneous analysis of multiple signals, while WT eliminates noise from acceleration response data, enhancing damage detection accuracy. The investigation reveals that the optimized SVMs, combined with MEMD and WT, significantly improve sensitivity, accuracy, and effectiveness over conventional methods. With an accuracy of 99.60%, the proposed method outperforms other approaches, which have an accuracy of 96.02%, 92.32% and lower for those not utilizing WT and MEMD or optimizing the parameters of SVM, respectively. This demonstrates superior predictive capabilities. Additionally, the choice of appropriate features extracted from the time-series data plays a crucial role in enhancing damage detection and assessing structural health conditions. By integrating these techniques, engineers are enabled to process large-scale data effectively, detect damage promptly, and make informed decisions to support civil infrastructure management practices.
Keywords
Main Subjects
)