Scholarlink

In this work, a unified framework integrating global and local SHM methods for structural health monitoring (SHM) of aerospace structures is proposed. This framework integrates both “local” ultrasonic-guided wave-based and “global” vibration-based SHM schemes for tackling damage detection, identification, and quantification under uncertainty. To achieve that, two modeling methods are applied including (i) the variational auto-encoder (VAE) latent space to represent the ultrasonic guided wave (GW) propagation, and (ii) auto-regressive models with exogenous excitation (ARX) models to capture low-frequency vibrations. Subsequent damage identification and quantification are performed based on a feed-forward neural network (FFNN) mapping the AE latent space representation and structural dynamics extracted by ARX models to the damage state. The complete experimental evaluation and assessment of the proposed framework are presented for an Airbus H125 helicopter blade under both low-frequency vibrations and ultrasonic guided-waves for SHM. Estimation results are compared with previous publications with single source of vibration or GW data.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Information Fusion of Ultrasonic Waves and Low-Frequency Vibrations: Leveraging Probabilistic Machine Learning and Stochastic Time Series Models for Structural Awareness

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Peiyuan Zhou,
  • Yiming Fan,
  • Fotis Kopsaftopoulos

摘要

In this work, a unified framework integrating global and local SHM methods for structural health monitoring (SHM) of aerospace structures is proposed. This framework integrates both “local” ultrasonic-guided wave-based and “global” vibration-based SHM schemes for tackling damage detection, identification, and quantification under uncertainty. To achieve that, two modeling methods are applied including (i) the variational auto-encoder (VAE) latent space to represent the ultrasonic guided wave (GW) propagation, and (ii) auto-regressive models with exogenous excitation (ARX) models to capture low-frequency vibrations. Subsequent damage identification and quantification are performed based on a feed-forward neural network (FFNN) mapping the AE latent space representation and structural dynamics extracted by ARX models to the damage state. The complete experimental evaluation and assessment of the proposed framework are presented for an Airbus H125 helicopter blade under both low-frequency vibrations and ultrasonic guided-waves for SHM. Estimation results are compared with previous publications with single source of vibration or GW data.