Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 11, 2026SensorsOpen Access

A Shapelet Transform-Based Method for Structural Damage Identification: A Case Study on a Wooden Truss Bridge

View Full Paper
Ask AI
Bookmark
Share

Authors

KGKe GanYYYalan YeFNFulin Nie

Discussion

Loading...

Member takes

Overview

Demonstrates a method that achieves high accuracy in identifying damage in wooden truss bridges, indicating improved interpretability and robustness.

Key Points

  • The aim is to develop a reliable method for identifying structural damage by overcoming limitations of current monitoring techniques.
  • Utilized measured random vibration response data from a timber truss bridge.
  • Applied Shapelet Transform to extract local subsequences with high information gain.
  • Classified extracted features using a Random Forest algorithm.
  • Investigated the impact of sensor locations, damage severities, and environmental variations.
  • Achieved 100% identification accuracy with a Shapelet extraction time of 10 minutes.
  • Average accuracies declined to 93.98%, 89.51%, and 58.48% with decreased extraction times.
  • Effectively identified minimum simulated damage of +23.5 g, only 0.07% of total mass.
  • Demonstrated robust performance despite variations in sensor locations and environmental conditions.

Cite This Study

Gan et al. (2026) studied this question.

synapsesocial.com/papers/69d9e67a78050d08c1b76dabhttps://doi.org/10.3390/s26082323
View Full Paper
Ask AI
Bookmark
Share