To address the challenge of difficult identification caused by the high coupling between leakage signals and load fluctuations in natural gas pipelines, an intelligent leakage identification method based on CNN (convolutional neural network) fused with SENet (squeeze-and-excitation network) was proposed. In this method, the spatiotemporal features of pressure and flow data was fused through a dual-input branch structure. The channel weights of leakage-sensitive features was dynamically enhanced by SENet to suppress load interference. And the data scale was compressed via GMM (Gaussian mixture model) clustering combined with orthogonal test design. Spatiotemporal thermal map analysis, ablation experiment analysis, and comparative test analysis were conducted based on the simulation data of three real pipeline networks in China. The results show that the proposed method exhibits excellent performance in tree-shaped, looped, and mixed topological structures, with an accuracy rate of 96.8%, a 3.4%~5.2% improvement compared with traditional leakage identification models. Furthermore, the sample size is reduced from tens of millions to thousands through data compression, which achieves a balance between accuracy and computational efficiency. The research results provide an intelligent solution for high-precision and low-latency monitoring of natural gas pipeline leakage.
DONG Chen-guang, HE Bing, WANG Mao-huai, LIU Si-yuan, WANG Zhao-jie, LU Xiao-qing.
Natural Gas Pipeline Leakage Identification Method Based on Convolutional Neural Network Fusion Squeeze-and-Excitation Network[J].
Science Technology and Engineering, 2026, 26(18): 7674-7682 DOI:10.12404/j.issn.1671-1815.2504710
GongJing, HaoYunqi, LuYangfan, et al. Application of artificial intelligence technology to the regulation and control of natural gas pipeline networks: technological innovation, evolutionary pathway, and strategic prospect[J]. Natural Gas Industry, 2025, 45(11): 203-216.
WangSuonan, ZhangLei, LiuShaohu. Safety inspection path planning for unmanned aerial vehicle in oil and gas fracturing unit[J]. Science Technology and Engineering, 2024, 24(25): 10832-10840.
FanJiarui, TangLiping, WangXijun, et al. Simulation on sensitivity of leakage diffusion parameters of buried gas pipeline[J]. Scie-nce Technology and Engineering, 2025, 25(25): 10671-10679.
[7]
WangK, LiC, JiaW, et al. Under-expanded jet and diffusion characteristics for small-hole leakage of hydrogen-blended natural gas in high-pressure pipelines[J]. Process Safety and Environmental Protection, 2024, 190: 195-211.
LinXiaohui, LiGang, YangWenming, et al. Simulation of buried natural gas pipeline leakage and diffusion process[J]. Science Technology and Engineering, 2025, 25(8): 3400-3414.
[10]
孙颖. 基于生成对抗网络的天然气管道泄技术研究[D]. 黑龙江: 东北石油大学, 2025.
[11]
SunYing. Research on gas pipeline leak detection technology based on generative adversarial network[D]. Heilongjiang: Northeast Petroleum University, 2021.
MengLingya, LiuCuiwei, LiuChao, et al. Characteristics extraction of acoustic leakage signal for natural gas pipelines[J]. Journal of China University of Petroleum(Natural Science Edition), 2014, 38(6): 153-160.
[14]
李禹琦. 基于VMD和特征融合的管道泄漏检测技术研究[D]. 大庆: 东北石油大学, 2024.
[15]
LiYuqi. Research on pipeline leak detection technology based on VMD and feature fusion[D]. Daqing: Northeast Petroleum University, 2024.
[16]
YeL, WangC, ZhouX, et al. Natural gas pipeline weak leakage detection based on negative pressure wave decomposition and feature enhancement[J]. Reliability Engineering & System Safety, 2025, 257: 110857.
ChenQiuyan, ZhangXinyan, HeMin, et al. Identification of pipeline thermal image leakage based on deep learning[J]. Infrared Technology, 2024, 46(5): 522-531.
ChenCen, JiYubo, WangHuan, et al. Gas pipeline network leak localization method based on transfer learning[J]. China Safety Science Journal, 2025, 35(3): 212-220.
[21]
MaH, ZhongY, WangJ, et al. Method for identifying the leakage of buried natural gas pipeline by soil vibration signals[J]. Gas Science and Engineering, 2024, 132: 205487.
LiBozhi, YangMinghe, XuKai, et al. Prediction method for ROP based on atention mechanism of convolutional neural network[J]. Science Technology and Engineering, 2024, 24(21): 8910-8916.
[24]
LiuR, ZhangL, JiaJ, et al. A 1DCNN-SENet-LSTM model for UGW-based automatic evaluation of grouting void location in post-tensioned concrete structure[J]. Engineering Structures, 2025, 339: 120669.
GaoXuke, QinChao, GaoFeng, et al. Dual parameter early warning model for winding faults in distribution dry-type transfor-mers based on LSTM-GMM[J]. Science Technology and Engineering, 2024, 24(31): 13407-13414.
[27]
JiangY, FuM, FangJ, et al. Advancing an LDA-GMM-CorEx topic model with prior domain knowledge in information systems research[J]. Information & Management, 2025, 62(2): 104097.