An Interpretable Multisensor Data Fusion Framework for Intelligent Monitoring of Historical Bridges

IF 5 2区 工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY
Quanjun Hua, Qing Chun, Yu Yuan, Xiaoyang Sun
{"title":"An Interpretable Multisensor Data Fusion Framework for Intelligent Monitoring of Historical Bridges","authors":"Quanjun Hua,&nbsp;Qing Chun,&nbsp;Yu Yuan,&nbsp;Xiaoyang Sun","doi":"10.1155/stc/5573568","DOIUrl":null,"url":null,"abstract":"<p>Historical bridges are often subjected to long-term environmental actions and gradual material degradation, while practical constraints on inspection and intervention make reliable condition assessment highly challenging. In this context, intelligent monitoring systems generate large volumes of heterogeneous sensor data, yet transforming such data into interpretable and actionable information for engineering decision-making remains a key problem. This study proposes an interpretable multisensor data fusion framework for intelligent monitoring of historical bridges, integrating temporal prediction and dependency modelling to achieve system-level anomaly assessment under strong environmental influence. Long short–term memory (LSTM) networks are employed to model the temporal evolution of individual sensor signals and to extract prediction residuals that represent deviations from expected behaviour. To capture nonlinear and non-Gaussian dependencies among heterogeneous sensors, a Copula-based dependency model is constructed on the residual space, enabling coordinated anomaly identification across structural and environmental measurements. A hierarchical fusion strategy is adopted to balance local sensor-group interactions and global system-level behaviour. Furthermore, Shapley value analysis is introduced to quantify the contribution of different sensor groups to detected anomalies, enhancing interpretability and supporting engineering judgement. The framework is validated using 34 months of continuous monitoring data from Lugou Bridge, a historical masonry arch bridge. The results demonstrate that system-level fusion effectively reduces scattered single-sensor alarms and highlights temporally concentrated anomalous periods consistent with documented extreme environmental events. The proposed framework is designed as a generalisable and interpretable solution for long-term multisensor monitoring, with the historical bridge serving as a representative case study.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0000,"publicationDate":"2026-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/5573568","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Structural Control & Health Monitoring","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1155/stc/5573568","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CONSTRUCTION & BUILDING TECHNOLOGY","Score":null,"Total":0}
引用次数: 0

Abstract

Historical bridges are often subjected to long-term environmental actions and gradual material degradation, while practical constraints on inspection and intervention make reliable condition assessment highly challenging. In this context, intelligent monitoring systems generate large volumes of heterogeneous sensor data, yet transforming such data into interpretable and actionable information for engineering decision-making remains a key problem. This study proposes an interpretable multisensor data fusion framework for intelligent monitoring of historical bridges, integrating temporal prediction and dependency modelling to achieve system-level anomaly assessment under strong environmental influence. Long short–term memory (LSTM) networks are employed to model the temporal evolution of individual sensor signals and to extract prediction residuals that represent deviations from expected behaviour. To capture nonlinear and non-Gaussian dependencies among heterogeneous sensors, a Copula-based dependency model is constructed on the residual space, enabling coordinated anomaly identification across structural and environmental measurements. A hierarchical fusion strategy is adopted to balance local sensor-group interactions and global system-level behaviour. Furthermore, Shapley value analysis is introduced to quantify the contribution of different sensor groups to detected anomalies, enhancing interpretability and supporting engineering judgement. The framework is validated using 34 months of continuous monitoring data from Lugou Bridge, a historical masonry arch bridge. The results demonstrate that system-level fusion effectively reduces scattered single-sensor alarms and highlights temporally concentrated anomalous periods consistent with documented extreme environmental events. The proposed framework is designed as a generalisable and interpretable solution for long-term multisensor monitoring, with the historical bridge serving as a representative case study.

历史桥梁智能监测的可解释多传感器数据融合框架
历史桥梁往往受到长期的环境作用和逐渐的材料退化,而检查和干预的实际限制使得可靠的状态评估极具挑战性。在这种情况下,智能监控系统会产生大量异构传感器数据,但将这些数据转化为可解释和可操作的信息以供工程决策仍然是一个关键问题。本研究提出了一个可解释的多传感器数据融合框架,用于历史桥梁智能监测,将时间预测和依赖建模相结合,实现强环境影响下的系统级异常评估。长短期记忆(LSTM)网络用于模拟单个传感器信号的时间演变,并提取代表偏离预期行为的预测残差。为了捕获异构传感器之间的非线性和非高斯依赖关系,在残差空间上构建了基于copula的依赖模型,实现了跨结构和环境测量的协调异常识别。采用分层融合策略平衡局部传感器组交互和全局系统级行为。此外,引入Shapley值分析来量化不同传感器组对检测异常的贡献,增强可解释性并支持工程判断。该框架通过对历史悠久的石拱桥卢沟桥34个月的连续监测数据进行验证。结果表明,系统级融合有效地减少了分散的单传感器报警,并突出了与记录的极端环境事件一致的时间集中异常周期。该框架被设计为一种通用的、可解释的长期多传感器监测解决方案,并以历史桥梁作为代表性案例研究。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Structural Control & Health Monitoring
Structural Control & Health Monitoring 工程技术-工程:土木
CiteScore
9.50
自引率
13.00%
发文量
234
审稿时长
8 months
期刊介绍: The Journal Structural Control and Health Monitoring encompasses all theoretical and technological aspects of structural control, structural health monitoring theory and smart materials and structures. The journal focuses on aerospace, civil, infrastructure and mechanical engineering applications. Original contributions based on analytical, computational and experimental methods are solicited in three main areas: monitoring, control, and smart materials and structures, covering subjects such as system identification, health monitoring, health diagnostics, multi-functional materials, signal processing, sensor technology, passive, active and semi active control schemes and implementations, shape memory alloys, piezoelectrics and mechatronics. Also of interest are actuator design, dynamic systems, dynamic stability, artificial intelligence tools, data acquisition, wireless communications, measurements, MEMS/NEMS sensors for local damage detection, optical fibre sensors for health monitoring, remote control of monitoring systems, sensor-logger combinations for mobile applications, corrosion sensors, scour indicators and experimental techniques.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信
小红书