Akademik Veri Yönetim Sistemi
ENGINEERING OPTIMIZATION, 2026 (SCI-Expanded, Scopus)
Laminated composite structures are widely used in engineering applications but remain prone to internal damage, such as delamination and matrix cracking. Vibration-based damage detection is an effective non-destructive approach, yet conventional optimization methods often suffer from high computational cost and premature convergence. This study introduces a golden ratio oppositional teaching-learning-based optimization (GROL-TLBO) algorithm, which integrates the golden ratio opposition-based learning mechanism into TLBO to enrich population diversity and accelerate convergence. The method formulates damage detection as an optimization problem using both natural frequencies and mode shapes, and is validated on laminated composite cantilever beams under multiple damage scenarios. Results show that GROL-TLBO requires only 50-100 iterations, compared to around 500 for TLBO, while maintaining prediction errors below 1% in noise-free and under 3% in noisy conditions, whereas TLBO errors exceed 5%, with false-positive detections. These findings highlight GROL-TLBO's superior accuracy, efficiency and robustness for practical structural health monitoring.