电器与能效管理技术 ›› 2024, Vol. 0 ›› Issue (8): 50-56.doi: 10.16628/j.cnki.2095-8188.2024.08.007

• 检测与装置 • 上一篇    下一篇

基于注意力机制优化电弧特征的光伏直流电弧故障检测方法

谢振华1,2,3, 刘玉莹4, 侯林明1,2,3, 王尧4, 周家旺4, 盛德杰4   

  1. 1.浙江省机电产品质量检测所有限公司, 浙江 杭州 310000
    2.智能电器试验与检测技术浙江省工程研究中心, 浙江 杭州 310051
    3.浙江省低压电器智能化与新能源应用重点实验室, 浙江 杭州 310051
    4.河北工业大学 省部共建电工装备可靠性与智能化国家重点实验室, 天津 300401
  • 收稿日期:2024-05-21 出版日期:2024-08-30 发布日期:2024-09-13
  • 作者简介:谢振华(1983—),男,高级工程师,主要从事低压电器检测技术的研究工作。|刘玉莹(1999—),女,硕士研究生,研究方向为电弧特征提取及识别技术。|侯林明(1995—),男,工程师,主要从事智能低压电器检测技术及检测装备的研究工作。
  • 基金资助:
    中央引导地方科技发展资金项目(226Z2102G);浙江省自然科学基金(LTGG23E070001);河北省高等学校科学技术研究项目(CXY2023006)

Photovoltaic DC Arc Fault Detection Method Based on Attention Mechanism Optimizing Arc Characteristics

XIE Zhenhua1,2,3, LIU Yuying4, HOU Linming1,2,3, WANG Yao4, ZHOU Jiawang4, SHENG Dejie4   

  1. 1. Zhejiang Testing & Inspection Institute for Mechanical and Electrical Products Quality Co.,Ltd., Hangzhou 310000, China
    2. Intelligent Electrical Testing and Testing Technology Zhejiang Engineering Research Center, Hangzhou 310051, China
    3. Key Laboratory of Low Voltage Apparatus Intelligentization and New Energy Application of Zhejiang Province, Hangzhou 310051, China
    4. State Key Laboratory of Reliability and Intelligence of Electrical Equipment,Hebei University of Technology, Tianjin 300401, China
  • Received:2024-05-21 Online:2024-08-30 Published:2024-09-13

摘要:

光伏系统中,因绝缘老化或接线松动而出现的直流串联电弧故障极易引发电气火灾。因此,光伏系统必须安装电弧故障检测装置,而其易因阴影遮挡和逆变器启动引发的直流侧高频噪声而误跳闸。将注意力机制和一维卷积神经网络相结合,提出一种基于注意力权重优化电弧特征的电弧故障检测方法。通过可视化电弧特征贡献权重,提取8~18 kHz和28~38 kHz电弧关键特征频段,并剔除8~23 kHz频段中的干扰特征频段。经验证,使用关键电弧特征训练的电弧故障检测模型可以成功避免阴影遮挡和逆变器启动过程带来的误动,最终使电弧检测准确率提高到99.33%。

关键词: 串联电弧故障, 卷积神经网络, 注意力机制, 电弧故障识别, 光伏系统

Abstract:

The DC series arc fault in photovoltaic systems caused by insulation aging or loose wiring is highly prone to electrical fires.Therefore,the arc fault detection devices must be installed in photovoltaic systems.However,the arc fault detection devices easily malfunction due to DC-side high-frequency noise caused by shadow occlusion and inverter startup.A novel arc fault detection method is proposed based on attention weight screening of arc features by combining the attention mechanism with the 1d convolutional neural network.By visualizing the contribution weight of arc features,the critical feature bands of 8~18 kHz and 28~38 kHz are extracted,and the interference arc features bands in the 8~23 kHz frequency band are removed.It has been verified that the arc fault detection model trained with the key arc features can successfully avoid the false activation caused by shadow occlusion and inverter startup,and the an arc detection accuracy of 99.33%is ultimately achieved.

Key words: series arc fault, convolutional neural network, attention mechanism, arc fault recognition, photovoltaic system

中图分类号: