Scholarlink

The increasing complexity of modern power systems and the growing reliance on smart grids demand, advanced fault identification and categorization techniques for power transmission lines (TLs) to ensure a stable and continuous electricity supply. Machine learning (ML) has become a potent technique for precisely and efficiently diagnosing faults. This paper presents a novel cascaded hybrid model combining quadratic discriminant analysis and random forest (QdDA-RF) to effectively identify and categorize faults in TL. The proposed model exhibits outstanding performance having accuracy of 99.81%; average values of precision, recall and F1 score are 99.83%, 99.83%, and 100% respectively. Moreover, the comparative analysis among various existing models reveals the proposed model’s effectiveness and robustness, making a significant advancement in TL fault diagnosis.

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

A Hybrid QdDA-RF Framework with Probabilistic Feature Augmentation for Power Transmission Line Fault Identification and Categorization

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mahmud Naeem,
  • Fariya Tabassum

摘要

The increasing complexity of modern power systems and the growing reliance on smart grids demand, advanced fault identification and categorization techniques for power transmission lines (TLs) to ensure a stable and continuous electricity supply. Machine learning (ML) has become a potent technique for precisely and efficiently diagnosing faults. This paper presents a novel cascaded hybrid model combining quadratic discriminant analysis and random forest (QdDA-RF) to effectively identify and categorize faults in TL. The proposed model exhibits outstanding performance having accuracy of 99.81%; average values of precision, recall and F1 score are 99.83%, 99.83%, and 100% respectively. Moreover, the comparative analysis among various existing models reveals the proposed model’s effectiveness and robustness, making a significant advancement in TL fault diagnosis.