摘要
为了解决极化-去极化电流法在目前海底电缆绝缘诊断中支路辨识易受噪声干扰的问题,基于信息熵理论提出一种改进的支路辨识定阶方法。首先分析导致传统支路辨识方法在噪声环境下定阶不够准确的理论原因,之后利用信息熵理论对奇异值所带信息进行阶数确定,最后分别对实验室新电缆以及水树老化电缆运用该方法进行绝缘支路辨识。结果表明:水树老化缺陷的存在导致电荷在电缆的去极化过程中的弛豫时间与其他极化形式不同,最终使得支路数增加。改进定阶法在对水树老化电缆进行支路辨识时,能够更为准确地辨识出水树老化对应的极化支路,验证了所提方法的有效性,并为海底电缆绝缘状态诊断提供参考。
In order to address the issue of susceptibility to noise interference in the branch identification of the polarization-depolarization current method in current submarine cable insulation diagnostics,an improved branch identification and order determination method based on information entropy theory is proposed.First,we analyze the theoretical reasons for the insufficient accuracy of traditional branch identification methods in noisy environments.Subsequently,using information entropy theory,we determine the order based on the information carried by singular values.Finally,we apply this method to insulation branch identification for both newly manufactured cables in the laboratory and aged cables with water tree defects.The results indicate that the presence of water tree aging defects leads to a different relaxation time of charges in the depolarization process compared to other polarization forms,ultimately resulting in an increase in the number of branches.When applying the improved order determination method to identify branches in water tree-aged cables,it accurately identifies the polarization branches corresponding to water tree aging,validating the effectiveness of the proposed approach and providing insights for submarine cable insulation state diagnostics.
作者
于是乎
林思衍
吴吉
陈祎林
余欣
周凯
陈宇慧
李银格
YU Shihu;LIN Siyan;WU Ji;CHEN Yilin;YU Xin;ZHOU Kai;CHEN Yuhui;LI Yinge(Electric Power Research Institute of Guangdong Power Grid Co.,Ltd.,Guangzhou,Guangdong 510080,China;College of Electrical Engineering,Sichuan University,Chengdu,Sichuan 610065,China)
出处
《广东电力》
2023年第11期130-137,共8页
Guangdong Electric Power
基金
中国南方电网有限责任公司科技项目(036100KK52200052(GDKJXM20202022))。
关键词
海底电缆
绝缘诊断
极化-去极化电流法
支路辨识
信息熵
submarine cable
insulation diagnosis
polarization and depolarization current method
branch identification
information entropy
