Scholarlink

Insulators are critical components in power transmission lines but can develop zero-value defects during long-term operation, threatening grid stability. Timely detection of these defects is essential for safe operation. This paper presents a non-contact method combining electric field simulation and interpolation prediction to identify zero-value defects in ceramic insulators. A finite element model was developed to analyze the effects of defect location and detection distance on spatial electric field distribution. Results showed distortion is most significant within 20–300 mm detection range, with double-plate continuous defects causing up to 37.4% distortion. To address data discreteness, a bilinear interpolation algorithm is introduced to construct a multi-condition prediction model and build an electric field feature database. The model achieves a prediction error below 1.13%, confirming its accuracy and adaptability. The proposed method provides effective support for intelligent maintenance and on-site zero-value defect identification in transmission systems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Study on Zero-Value Insulator Identification Based on Electric Field Simulation and Interpolation Prediction

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ruifang Ding,
  • Hao Yang,
  • Yuanshuai He,
  • Yangzhen Zhao

摘要

Insulators are critical components in power transmission lines but can develop zero-value defects during long-term operation, threatening grid stability. Timely detection of these defects is essential for safe operation. This paper presents a non-contact method combining electric field simulation and interpolation prediction to identify zero-value defects in ceramic insulators. A finite element model was developed to analyze the effects of defect location and detection distance on spatial electric field distribution. Results showed distortion is most significant within 20–300 mm detection range, with double-plate continuous defects causing up to 37.4% distortion. To address data discreteness, a bilinear interpolation algorithm is introduced to construct a multi-condition prediction model and build an electric field feature database. The model achieves a prediction error below 1.13%, confirming its accuracy and adaptability. The proposed method provides effective support for intelligent maintenance and on-site zero-value defect identification in transmission systems.