A damage identification method for cable-stayed bridge multi-source monitoring data fusion

By synchronously collecting acceleration and displacement sensor data in cable-stayed bridges, constructing Gram matrix and CNN-LSTM model, the fusion of multi-source data is realized, solving the problems of data singularity and insufficient spatial correlation in bridge damage identification, improving identification accuracy and reliability, and making it suitable for health monitoring of cable-stayed bridges.

CN122196723APending Publication Date: 2026-06-12HANGZHOU POLYTECHNIC +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HANGZHOU POLYTECHNIC
Filing Date
2026-01-28
Publication Date
2026-06-12

AI Technical Summary

Technical Problem

Existing bridge damage identification technologies suffer from problems such as data uniformity, insufficient spatial correlation, and low identification accuracy, making it difficult to meet the high-precision engineering requirements.

Method used

A multi-source monitoring data fusion method for cable-stayed bridges is adopted. Through a spatial-temporal dual-channel fusion mechanism, data is collected synchronously using accelerometers and displacement sensors. A Gram matrix is ​​constructed to extract global spatial features, and dynamic features are extracted by combining the CNN-LSTM temporal module. The resulting joint features are then input into a classification and recognition model for damage identification.

Benefits of technology

It improves the comprehensiveness, stability and accuracy of damage identification, enabling early identification of minor damage and providing reliable basis for health monitoring and maintenance decisions. It is applicable to cable-stayed bridges and other long-span bridge structures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122196723A_ABST
    Figure CN122196723A_ABST
Patent Text Reader

Abstract

The application discloses a kind of cable-stayed bridge multi-source monitoring data fusion damage identification methods, including cable-stayed bridge sensor arrangement and multi-channel synchronous acquisition;Gram matrix is constructed to extract global spatial features, while CNN-LSTM time sequence module modeling extracts dynamic characteristics;Global spatial features and dynamic characteristics are fused to form joint features;Joint features are input into classification and identification model, and no damage, mild damage, moderate damage results are output based on damage degree based on damage degree, and through the damage identification method, space-time dual-channel fusion mechanism can grasp the collaborative response mode of structure at the global level, track the damage evolution process at the time level, improve the comprehensiveness, stability and accuracy of damage identification, and provide reliable basis for cable-stayed bridge health monitoring and maintenance decision.
Need to check novelty before this filing date? Find Prior Art