Scholarlink

<p>Fault detection techniques have been widely used in many industrial systems over the past few decades. With the rapid development of railway electrification worldwide, electric traction drive has become an indispensable part of modern rail transit systems. The complex operating condition of the electrical traction drive system causes ground fault, which change the key components of railway traction drive system. Therefore, effective fault diagnosis method in railway traction system is proposed in this work. In the proposed technique, Empirical Wavelet Transform (EWT) is used to extract the features from railway traction system before and after applying the faults. Hybrid artificial neural network (ANN) and bidirectional long short-term memory (Bi-LSTM) is used to classify the defect that has occurred. The proposed method is implemented in the MATLAB and Simulink tool and the results are evaluated in terms of different performance metrics. Further, the performance of the proposed fault detection method is related with recent prevailing methods. Simulation validation demonstrate that the proposed method achieves 99.72% accuracy with a Mean Absolute Error (MAE) of 0.01%, outperforming conventional classifiers in terms of accuracy, robustness, and computational efficiency. The results confirm that the proposed framework is well-suited for real-time fault detection and predictive maintenance in modern railway traction systems.</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

Intelligent Fault Diagnosis in Electric Railway Traction System: A Novel Hybrid Approach

  • 期刊论文

  • 发表日期：

    2026年3月2日

查看全文
  • Rohini Sharma

摘要

Fault detection techniques have been widely used in many industrial systems over the past few decades. With the rapid development of railway electrification worldwide, electric traction drive has become an indispensable part of modern rail transit systems. The complex operating condition of the electrical traction drive system causes ground fault, which change the key components of railway traction drive system. Therefore, effective fault diagnosis method in railway traction system is proposed in this work. In the proposed technique, Empirical Wavelet Transform (EWT) is used to extract the features from railway traction system before and after applying the faults. Hybrid artificial neural network (ANN) and bidirectional long short-term memory (Bi-LSTM) is used to classify the defect that has occurred. The proposed method is implemented in the MATLAB and Simulink tool and the results are evaluated in terms of different performance metrics. Further, the performance of the proposed fault detection method is related with recent prevailing methods. Simulation validation demonstrate that the proposed method achieves 99.72% accuracy with a Mean Absolute Error (MAE) of 0.01%, outperforming conventional classifiers in terms of accuracy, robustness, and computational efficiency. The results confirm that the proposed framework is well-suited for real-time fault detection and predictive maintenance in modern railway traction systems.