基于多模态融合的长输管道安全监测研究（特邀）

朱程远; 蒲岩云; 杨楷翔; 杨秦敏

doi:10.13756/j.gtxyj.2026.260124

您当前的位置：

首页 >

文章列表页 >

基于多模态融合的长输管道安全监测研究（特邀）

专题：2025全国硕博士光通信菁英论坛 | 更新时间：2026-10-21

- 基于多模态融合的长输管道安全监测研究（特邀）
- Research on Safety Monitoring of Long-Distance Pipelines based on Multimodal Fusion
- 光通信研究 2026年第5期页码：22-32
- 作者机构：
  
  1.浙江大学　控制科学与工程学院　工业控制技术全国重点实验室，杭州　310027
  2.华南理工大学　计算机科学与工程学院，广州　510006
  3.浙江省全省数智管理与决策技术重点实验室，杭州　310058
- 作者简介：
  
  朱程远（1996-），男，山东济宁人。博士，主要研究方向为分布式光纤传感。
  杨秦敏，教授。E-mail：qmyang@zju.edu.cn
- 基金信息：
  
  浙江省自然科学基金白马湖实验室联合基金资助项目(LBMHD24F030001);长三角科技创新共同体联合攻关重点揭榜任务资助项目(20259b09050007);工业控制技术全国重点实验室浙大专项资助项目(ICT2026C05)
- DOI：10.13756/j.gtxyj.2026.260124
  中图分类号： TN929
- 收稿：2026-05-05，
  
  修回：2026-07-15，
  
  纸质出版：2026-10-10
- 稿件说明：
移动端阅览
朱程远，蒲岩云，杨楷翔，等. 基于多模态融合的长输管道外部安全监测研究[J]. 光通信研究，2026(5): 22-32.

Zhu C Y, Pu Y Y, Yang K X, et al. Research on Safety Monitoring of Long-Distance Pipelines based on Multimodal Fusion[J]. Study on Optical Communications, 2026(5): 22-32.
朱程远，蒲岩云，杨楷翔，等. 基于多模态融合的长输管道外部安全监测研究[J]. 光通信研究，2026(5): 22-32. DOI： 10.13756/j.gtxyj.2026.260124.

Zhu C Y, Pu Y Y, Yang K X, et al. Research on Safety Monitoring of Long-Distance Pipelines based on Multimodal Fusion[J]. Study on Optical Communications, 2026(5): 22-32. DOI： 10.13756/j.gtxyj.2026.260124.

摘要

目的

针对长输管道复杂场景下外部振动信号类间相似、类内波动明显以及单一模态特征难以全面表征事件差异的问题，文章提出了一种基于多模态融合的管道外部安全监测方法，以提高分布式声学传感（DAS）系统在实际运行环境中的事件识别性能。

方法

文章所提方法通过省级智慧管网平台DAS系统采集真实场景振动信号，将信号分别构建为原始波形、短时傅里叶变换时频图和格拉姆角场图像3种模态。针对不同模态的数据表达特点，构建多分支特征提取网络，分别利用一维卷积神经网络提取局部时序特征，利用二维卷积神经网络提取时频纹理与能量分布特征，利用视觉Transformer提取图像编码中的全局相关结构特征。在此基础上，设计跨模态注意力融合机制，对不同模态特征进行对齐、交互和融合；同时考虑实际场景中的类别不平衡与模态融合偏差问题，构建跨模态协同焦点损失函数，以增强模型对难分类样本和少数类样本的学习能力。

结果

真实管道外部事件识别数据集上的实验结果表明，文章所提方法训练过程收敛稳定，在管道外部事件识别任务中取得了较好的综合性能，平均精确率、平均召回率和平均F1分数分别达到0.968 3、0.972 8和0.970 5，均优于对比模型；单模态与多模态对比实验进一步表明，多模态输入能够提供更充分的判别信息，明显提升了模型对复杂振动信号的表征能力。

结论

文章所提方法能够有效融合原始时序、时频分布和全局相关结构信息，增强复杂背景下相似事件的可分性，为DAS管道外部安全监测与智能预警提供了一种有效的技术方案。

Abstract

Objective

To address the problems of inter-class similarity

significant intra-class variation

and insufficient representation capability of single-modal features for external vibration signals in complex long-distance oil and gas pipeline scenarios

a multimodal fusion-based pipeline external safety monitoring method is proposed to improve the event recognition performance of Distributed Acoustic Sensing (DAS) systems in practical operating environments.

Methods

Real-world vibration signals were collected by a DAS system deployed on a provincial smart pipeline network platform. Each signal sample was constructed into three modalities

namely raw waveform

short-time Fourier transform spectrogram

and Gramian angular field image. According to the data representation characteristics of different modalities

a multi-branch feature extraction network was developed. A one-dimensional convolutional neural network was used to extract the local temporal features. A two-dimensional convolutional neural network was used to extract time-frequency texture and energy distribution features

and a vision Transformer was used to extract global correlation structure features from encoded images. On this basis

a cross-modal attention fusion mechanism was designed to align

interact

and fuse heterogeneous modal features. Meanwhile

considering class imbalance and modal fusion bias in real-world scenarios

a cross-modal collaborative focal loss function was constructed to enhance the learning capability of the model for hard-to-classify and minority-class samples.

Results

Experimental results on a real pipeline external event recognition dataset show that the proposed method achieves stable convergence during training and good overall performance in the pipeline external event recognition task. The average precision

average recall

and average F1-score reach 0.968 3

0.972 8

and 0.970 5

respectively

outperforming the comparison models. In addition

the comparison between single-modal and multimodal inputs further demonstrates that multimodal input provides more sufficient discriminative information and significantly improves the representation capability of the model for complex vibration signals.

Conclusion

The proposed method can effectively integrate raw temporal information

time-frequency distribution

and global correlation structure

thereby improving the separability of similar events under complex backgrounds. It provides an effective technical solution for DAS-based pipeline external safety monitoring and intelligent early warning.

关键词

Keywords

references

Zhu C , Pu Y , Yang K , et al . Distributed Optical Fiber Intrusion Detection by Image Encoding and SwinT in Multi-Interference Environment of Long-Distance Pipeline [J ] . IEEE Transactions on Instrumentation and Measurement , 2023 , 72 : 2515012 .

Wu H , Chen J , Liu X , et al . One-Dimensional CNN-based Intelligent Recognition of Vibrations in Pipeline Monitoring with DAS [J ] . Journal of Lightwave Technology , 2019 , 37 ( 17 ): 4359 - 4366 .

王龙生 , 谢欣 , 姚猛 , 等 . 分布式光纤声波传感及其进展研究 [J ] . 光通信研究 , 2025 ( 5 ): 240100 .

Wang L S , Xie X , Yao M , et al . Distributed Optical Fiber Acoustic Wave Sensing and Its Progress [J ] . Study on Optical Communications , 2025 ( 5 ): 240100 .

Zhu C , Yang Y , Yang K , et al . Monitoring Landslide Disturbances Using Distributed Acoustic Sensing under Extreme Weather Conditions [J ] . npj Natural Hazards , 2026 , 3 : 23 .

Wu D , Liang Q Q , Hu B X , et al . Research Progress of Event Intelligent Perception based on DAS [J ] . Sensors , 2025 , 25 ( 16 ): 5052 .

丁美琪 , 桂林 , 王子怡 , 等 . 基于1D-CNN的射频强度温度传感研究 [J ] . 光通信研究 , 2025 ( 2 ): 240159 .

Ding M Q , Gui L , Wang Z Y , et al . Research on RF Intensity Temperature Sensing based on 1D-CNN [J ] . Study on Optical Communications , 2025 ( 2 ): 240159 .

Zhu C , Pu Y , Lyu Z , et al . A Lightweight Pipeline Edge Detection Model based on Heterogeneous Knowledge Distillation [J ] . IEEE Transactions on Circuits and Systems II: Express Briefs , 2024 , 71 ( 12 ): 5059 - 5063 .

何欣 . 基于分布式光纤振动传感的长输埋地管道安全监测技术 [J ] . 山西建筑 , 2020 , 46 ( 7 ): 197 - 198 .

He X . Safety Monitoring of Long-Distance Buried Pipeline based on Distributed Vibration Fiber Optical Fiber Sensor [J ] . Shanxi Architecture , 2020 , 46 ( 7 ): 197 - 198 .

Zhou X , Wu Z . Research Progress of Pipeline Health Monitoring based on Distributed Fiber Optic Sensing [C ] // 2024 9th International Conference on Electronic Technology and Information Science (ICETIS) . Hangzhou, China : IEEE , 2024 : 10593770 .

Zhu C , Pu Y , Lyu Z , et al . Multidimensional Information Fusion and Broad Learning System-based Condition Recognition for Energy Pipeline Safety [J ] . Knowledge-Based Systems , 2024 , 300 : 112259 .

Nie R , Sun Y , Wang S . Research on Underground Pipeline Illegal Intrusion Monitoring based on Distributed Optical Fiber Sensing Technology [C ] // First Conference on Distributed Optical Fiber Sensing Technology and Applications (DOFS2024) . Nanjing, China : SPIE , 2025 : 136540C .

张均伟 , 蓝波 , 黄嘉庚 , 等 . 基于分布式光纤传感的光缆防外破监测研究 [J ] . 光通信研究 , 2022 , ( 5 ): 53 - 57 .

Zhang J W , Lan B , Huang J Q , et al . Monitoring and Early Warning System for Anti-breakage of Optical Cable based on Distributed Optical Fiber Vibration Sensing [J ] . Study on Optical Communications , 2022 , ( 5 ): 53 - 57 .

Wen C , Hu S , Wang J , et al . Fiber Optic Sensing Technology in Underground Pipeline Health Monitoring: a Comprehensive Review [J ] . Structural Health Monitoring , 2026 , 25 ( 1 ): 630 - 658 .

Zhu C , Yang Y , Yang K , et al . AI-based Energy Transportation Safety: Pipeline Radial Threat Estimation Using Intelligent Sensing System [J ] . Proceedings of the AAAI Conference on Artificial Intelligence , 2024 , 38 ( 20 ): 22556 - 22564 .

Zhu C , Pu Y , Yang K , et al . A Novel Visual Transformer for Long-Distance Pipeline Pattern Recognition in Complex Environment [J ] . IEEE Transactions on Artificial Intelligence , 2024 , 5 ( 6 ): 2933 - 2945 .

浏览量

下载量

CSCD

CNKI被引量

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于分布式光纤传感的光缆防外破监测研究

基于改进EfficientNet-B0的φ-OTDR振动事件识别方法

基于分类器置信度的DAS信号定位和识别技术

基于VMD-GAF-MCNN的多传感融合海缆裸露状态识别

通感一体化光网络的技术、应用与展望