Scholarlink

<p>Ensuring the stability and reliability of power transmission networks requires accurate and timely fault detection. Conventional fault detection methods often struggle with adaptability, computational efficiency, and robustness to noisy conditions. This work proposes an intelligent transmission line fault detection and classification through an advanced deep learning based model. The proposed method converts time series current and voltage signals into structured feature representations by using fast Markov transition field transformation, and processes these feature representations with a ResNet34 based deep feature extractor with squeeze and excitation module. The proposed model enhances classification accuracy while decreasing the computational complexity, enabling real time fault detection. A 735&#xa0;kV transmission line system is modeled in the Simulink environment to generate a diverse dataset under various fault scenarios. The proposed model can detect such faults with an accuracy of 99.98%, which is higher than the state of the art fault detection techniques. The statistical significance of the model’s performance superiority is confirmed by the comparative analysis using the Wilcoxon signed-rank test. Further, the robustness of the model is validated against Gaussian noise, missing data and outliers, and the model is shown to be reliable in real world transmission systems. The results show that the proposed model is a computationally efficient, accurate and interpretable solution for modern power system protection.</p>

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

An intelligent approach for detection and classification of transmission line faults using Markov transition field and residual network

  • 期刊论文

  • 发表日期：

    2026年1月24日

查看全文
  • V. Pradeep,
  • K. Baskaran,
  • S. Ida Evangeline

摘要

Ensuring the stability and reliability of power transmission networks requires accurate and timely fault detection. Conventional fault detection methods often struggle with adaptability, computational efficiency, and robustness to noisy conditions. This work proposes an intelligent transmission line fault detection and classification through an advanced deep learning based model. The proposed method converts time series current and voltage signals into structured feature representations by using fast Markov transition field transformation, and processes these feature representations with a ResNet34 based deep feature extractor with squeeze and excitation module. The proposed model enhances classification accuracy while decreasing the computational complexity, enabling real time fault detection. A 735 kV transmission line system is modeled in the Simulink environment to generate a diverse dataset under various fault scenarios. The proposed model can detect such faults with an accuracy of 99.98%, which is higher than the state of the art fault detection techniques. The statistical significance of the model’s performance superiority is confirmed by the comparative analysis using the Wilcoxon signed-rank test. Further, the robustness of the model is validated against Gaussian noise, missing data and outliers, and the model is shown to be reliable in real world transmission systems. The results show that the proposed model is a computationally efficient, accurate and interpretable solution for modern power system protection.