College of Electrical and Control Engineering, Xi'an University of Science and Technology, Xi'an 710054, China
Show less
文章历史+
Received
Published
Revised
2025-08-08
2026-05-18
2025-11-26
Online Date
2026-08-06
PDF (6582K)
摘要
针对当前电气火灾预警多依赖明显故障特征、难以在风险萌芽阶段及时响应。聚焦于电气线路超温发生前的早期征兆阶段。提出了一种融合双层长短期记忆网络(long short-term memory, LSTM)增强型自编码器(autoencoder, AE)与单分类支持向量机(one-class support vector machine, OCSVM)的电气线路超温早期征兆异常识别方法。通过分析过载、谐波和非周期电流三类电气线路超温早期征兆的致灾原理与波形。构建数据集。利用AE-2LSTM结构有效学习电流数据的深层时序特征并实现高精度重构,在此基础上,利用正常与异常样本在重构误差上的显著差异,通过引入OCSVM模块实现仅依赖正常样本的单分类异常检测。实验结果表明,该方法识别准确率达98.16%,克服了电气火灾监测技术研究中异常数据获取困难、正负样本极度不平衡等实际建模难题,为电气火灾的早期预警与主动防控提供了重要的科学依据和技术支持。
Abstract
Currently, the early warning of electrical fires primarily relies on evident fault characteristics, making it challenging to promptly respond during the initial stages of risk development. The pre-overheating symptom phase in electrical circuits was focused on, and an enhanced AE (autoencoder) framework that integrates a two-layer LSTM (long short-term memory network) with OCSVM (one-class support vector machine) was proposed for detecting early signs of circuit overheating.By analyzing the underlying disaster mechanisms and current waveforms associated with three types of circuit conditions—overload, harmonic distortion, and aperiodic currents—a comprehensive dataset was constructed. The AE-2LSTM architecture was employed to effectively capture deep temporal features in current time series data, enabling high-precision signal reconstruction. Leveraging the significant disparity in reconstruction errors between normal and anomalous samples, the OCSVM module was introduced to perform single-class anomaly detection using only normal training samples. Experimental results demonstrate that the proposed method achieves an anomaly recognition accuracy of 98.16%, effectively addressing practical challenges in electrical fire monitoring, such as the scarcity of abnormal data and severe class imbalance. This approach provides a robust scientific foundation and technical support for the early warning, proactive prevention, and control of electrical fires.
PAN Hong-guang, CHEN Jia-tao, GUO Qiang, GU Tian-yu.
Abnormal Identification Method for Early Signs of Electrical Circuit Overheating Based on AE-2LSTM-OCSVM[J].
Science Technology and Engineering, 2026, 26(14): 6015-6023 DOI:10.12404/j.issn.1671-1815.2505825
JiaYizhuo, LiChangzheng, WangJingwu, et al. The influence of wire length on the burning behavior of overcurrent polyvinyl chloride copper wire[J]. Fire Science and Technology, 2024, 43(4): 572-577.
ZhouWei. Research on fire investigation and prevention technology of distributed photovoltaic power generation systems[J]. Fire Science and Technology, 2025, 44(6): 869-876.
[7]
NazirA, MoslehH, TakruriM, et al. Early fire detection: a new indoor laboratory dataset and data distribution analysis[J]. Fire-Switzerland, 2022, 5(1): 92-101.
LeiFangfei, ChuJifeng, LiuYang, et al. Cable overheating fault diagnosis method based on semiconductor gas sensing array[J]. Transactions of China Electrotechnical Society, 2023, 38(13): 3651-3664.
[10]
WangY W, LiQ H, ZhangJ Y, et al. A gas detection system combined with a global extension extreme learning machine for early warning of electrical fires[J]. Sensors and Actuators B-Chemical, 2025, 423(7): 136-145.
LiPeizhi, LiuLiqun, BoYishui. Abnormal electricity consumption detection based on density subspace isolated forest[J]. Science Technology and Engineering, 2024, 24(10): 4115-4123.
YangLei, GuoPeng, ZhangYuxiao. Abnormal data recognition method for wind turbine units based on manifold learning[J]. Distributed Energy, 2025, 11(1): 11-19.
ZhongWujun, LiPeiqiang, TuChunming, et al. Wind power anomaly data processing based on combined anomaly recognition and GAT-GAN[J]. Power System Automation, 2025, 49(12): 1-21.
XuXiaonan, NingXin, ZhuYuanpeng, et al. Application of fuzzy comprehensive evaluation-BP neural network in electrical fire risk assessment[J] Journal of Safety and Environment, 2025, 25(1): 1-10.
PanMingming, WangBaigen, QiHongtao, et al. Multimodal fusion model for early fire detection of electrical facilities[J]. Application of Electronic Technology, 2025, 51(6): 10-15.
MenMaochen, DuYujia, XuMingming, et al. Research on intelligent electrical fire monitoring based on residual current waveform characteristic quantities[J]. Fire Science and Technology, 2023, 42(4): 549-554.
[23]
KangY, JangE, ImJ, et al. A deep learning model using geostationary satellite data for forest fire detection with reduced detection latency[J]. Giscience & Remote Sensing, 2022, 59(1): 2019-2035.
HeSheng, ShuXueming, HuJun, et al. Electrical fire risk prediction and early warning method based on fire big data[J]. Journal of Tsinghua University (Natural Science Edition), 2024, 64(3): 478-491.
[26]
ZhangY, ChenH C, LiZ, et al. Lightweight AC arc fault detection method by integration of event-based load classification[J]. IEEE Transactions on Industrial Electronics, 2024, 7(4): 4130-4140.
SongZhenyu, XingChen, LiMing, et al. Fault diagnosis and fire risk assessment technology for photovoltaic DC distribution[J]. Global Energy Interconnection, 2024, 7(1): 46-54.
[29]
LiuH, LiJ C, WangW J, et al. Arc-1DCNN: an enhanced model for series arc fault detection[J]. Measurement, 2024, 234(4): 114-125.
ChenXiaolong, HuangGuozhong, GaoXuehong, et al. Research on temperature rise of single-core PVC cables in overload, damage and aging risk scenarios[J]. China Safety Production Science and Technology, 2022, 18(6): 25-30.
[32]
HuoY, XuX Y, LiY, et al. Experimental and theoretical research on the temperature evolution law of overcurrent fault wires[J]. Fire and Materials, 2024, 48(6): 632-641.
LiuSude, WangZixin, JingZiyang, et al. A review on electrical fires caused by arc faults in aviation cables[J]. Journal of Electrical Engineering, 2023, 18(3): 18-34.
[35]
MiaoW C, WangZ F, WangF, et al. Multicharacteristics arc model and autocorrelation-algorithm based arc fault detector for DC microgrid[J]. IEEE Transactions on Industrial Electronics, 2023, 70(5): 4875-4886.
SongWei, LinJianwei, ZhouFangze, et al. Fault diagnosis method for fan bearings based on improved noise reduction autoencoder[J]. Power System Protection and Control, 2022, 50(10): 61-68.
[38]
LiX C, XieK, WangX, et al. A light-weight and robust tensor convolutional autoencoder for anomaly detection[J]. IEEE Transactions on Knowledge and Data Engineering, 2024, 36(9): 4346-4360.
LiYing, HaoJian, DingYilin, et al. Mechanical defect diagnosis method for gis equipment based on multimodal combined vibration images and stacked sparse autoencoders[J]. High Voltage Technology, 2025, 51(2): 753-765.
LuQirong, WangZexin, YeYingya, et al. Load clustering based on VMD and spatio-temporal network variational autoencoder[J]. Science Technology and Engineering, 2024, 24(14): 5831-5838.
GongXuanlin, TaoQing, SuNa, et al. Recognition method of steady-state visual evoked potential electroencephalogram signals based on CNN-CBAM-LSTM[J]. Science Technology and Engineering, 2025, 25(10): 4175-4182.