Intelligent Fault Diagnosis of the High-Speed Train With Big Data Based on Deep Neural Networks

A wheel set bearing is an important supporting component of a high-speed train. Its quality and performance directly determine the overall safety of the train. Therefore, monitoring a wheel set bearing’s conditions for an early fault diagnosis is vital to ensure the safe operation of high-speed trains. However, the collected signals are often contaminated by environmental noise, transmission path, and signal attenuation because of the complexity of high-speed train systems and poor operation conditions, making it difficult to extract the early fault features of the wheel set bearing accurately. Vibration monitoring is most widely used for bearing fault diagnosis, with the acoustic emission (AE) technology emerging as a powerful tool. This article reports a comparison between vibration and AE technology in terms of their applicability for diagnosing naturally degraded wheel set bearings. In addition, a novel fault diagnosis method based on the optimized maximum second-order cyclostationarity blind deconvolution (CYCBD) and chirp Z-transform (CZT) is proposed to diagnose early composite fault defects in a wheel set bearing. The optimization CYCBD is adopted to enhance the fault-induced impact response and eliminate the interference of environmental noise, transmission path, and signal attenuation. CZT is used to improve the frequency resolution and match the fault features accurately under a limited data length condition. Moreover, the efficiency of the proposed method is verified by the simulated bearing signal and the real datasets. The results show that the proposed method is effective in the detection of wheel set bearing faults compared with the minimum entropy deconvolution (MED) and maximum correlated kurtosis deconvolution (MCKD) methods. This research is also the first to compare the effectiveness of applying AE and vibration technologies to diagnose a naturally degraded high-speed train bearing, particularly close to actual line operation conditions.

Download Full-text

Enabling deeper learning on big data for materials informatics applications

Scientific Reports ◽

10.1038/s41598-021-83193-1 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Dipendra Jha ◽

Vishu Gupta ◽

Logan Ward ◽

Zijiang Yang ◽

Christopher Wolverton ◽

...

Keyword(s):

Neural Networks ◽

Big Data ◽

Deep Learning ◽

Deep Neural Networks ◽

Materials Science ◽

Prediction Models ◽

Model Performance ◽

Materials Informatics ◽

Learning Framework ◽

Significant Attention

AbstractThe application of machine learning (ML) techniques in materials science has attracted significant attention in recent years, due to their impressive ability to efficiently extract data-driven linkages from various input materials representations to their output properties. While the application of traditional ML techniques has become quite ubiquitous, there have been limited applications of more advanced deep learning (DL) techniques, primarily because big materials datasets are relatively rare. Given the demonstrated potential and advantages of DL and the increasing availability of big materials datasets, it is attractive to go for deeper neural networks in a bid to boost model performance, but in reality, it leads to performance degradation due to the vanishing gradient problem. In this paper, we address the question of how to enable deeper learning for cases where big materials data is available. Here, we present a general deep learning framework based on Individual Residual learning (IRNet) composed of very deep neural networks that can work with any vector-based materials representation as input to build accurate property prediction models. We find that the proposed IRNet models can not only successfully alleviate the vanishing gradient problem and enable deeper learning, but also lead to significantly (up to 47%) better model accuracy as compared to plain deep neural networks and traditional ML techniques for a given input materials representation in the presence of big data.

Download Full-text