Scholarlink

Predicting the aging state of high-voltage electrical equipment is critical to preventing failures that may result in safety hazards and substantial economic losses. This work introduces a differential measurement approach that mitigates system-specific bias, thereby improving the generalization of machine learning models to new apparatuses. Experimental validation on real-world data shows a 38% reduction in mean absolute error on unseen test cases, highlighting the practical value of the method.

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

Automatic Aging Prediction of High Voltage Apparatus Based on Differential Measurements

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Edoardo Ragusa,
  • Christian Gianoglio,
  • Laura Della Giovanna

摘要

Predicting the aging state of high-voltage electrical equipment is critical to preventing failures that may result in safety hazards and substantial economic losses. This work introduces a differential measurement approach that mitigates system-specific bias, thereby improving the generalization of machine learning models to new apparatuses. Experimental validation on real-world data shows a 38% reduction in mean absolute error on unseen test cases, highlighting the practical value of the method.