Search Results (5)

Diagnosing faults in wheelset bearings is critical for train safety. The main challenge is that only a limited amount of fault sample data can be obtained during high-speed train operations. This scarcity of samples impacts the training and accuracy of deep learning models for wheelset bearing fault diagnosis. Studies show that the Auxiliary Classifier Generative Adversarial Network (ACGAN) demonstrates promising performance in addressing this issue. However, existing ACGAN models have drawbacks such as complexity, high computational expenses, mode collapse, and vanishing gradients. Aiming to address these issues, this paper presents the Transformer and Auxiliary Classifier Generative Adversarial Network (TACGAN), which increases the diversity, complexity and entropy of generated samples, and maximizes the entropy of the generated samples. The transformer network replaces traditional convolutional neural networks (CNNs), avoiding iterative and convolutional structures, thereby reducing computational expenses. Moreover, an independent classifier is integrated to prevent the coupling problem, where the discriminator is simultaneously identified and classified in the ACGAN. Finally, the Wasserstein distance is employed in the loss function to mitigate mode collapse and vanishing gradients. Experimental results using the train wheelset bearing datasets demonstrate the accuracy and effectiveness of the TACGAN. Full article

The bogie frame, as one of the most critical load-bearing structures of the Electric Multiple Unit (EMU), is responsible for bearing and transmitting various loads from the car body, wheelsets, and its own installation components. With the increasing operating speed of high-speed EMUs, especially when the design and operational speeds exceed 400 km/h, the applicability of current international standards is uncertain. The load spectrum serves as the foundation for structural reliability design and fatigue evaluation. In this paper, the measured loads of the bogie frame of a CR400AF high-speed train on the Beijing–Shanghai high-speed railway are obtained, and the time-domain characteristic of the measured loads is analyzed under different operating conditions. Then, through the Weibull distribution of three parameters, the Weibull parameters at the 450 km/h speed level are fitted, and the maximum load and cumulative frequency under the speed level are derived. Finally, the load spectrum of the bogie frame at the 450 km/h speed level is established, which provides a more realistic load condition for accurately evaluating the fatigue strength of bogie frames at higher speed levels. Full article

Vibration signal analysis is an important means for bearing fault diagnosis. Affected by the vibration of other machine parts, external noise and the vibration transmission path, the impulses induced by a bearing defect in the measured vibrations are very weak. Blind deconvolution (BD) methods can counteract the effect of the transmission path and enhance the fault impulses. Most BD methods highlight fault features of the filtered signals by impulse-featured objective functions (OFs). However, residual noise in the filtered signals has not been well tackled. To overcome this problem, a fuzzy entropy-assisted deconvolution (FEAD) method is proposed. First, FEAD takes advantage of the high noise sensitivity of fuzzy entropy (FuzzyEn) and constructs a weighted FuzzyEn–kurtosis OF to enhance the fault impulses while suppressing noise interference. Then, the PSO algorithm is used to iteratively solve the optimal inverse deconvolution filter. Finally, envelope spectrum analysis is performed on the filtered signal to realize bearing fault diagnosis. The feasibility of FEAD was first verified by the bearing fault simulation signals at constant and variable speeds. The bearing test signals from Case Western Reserve University (CWRU), the railway wheelset and the test bench validated the good performance of FEAD in fault feature enhancement. A comparison with and quantitative results for the other state-of-the-art BD methods indicated the superiority of the proposed method. Full article

A wheelset bearing is one of the main components of the train bogie frame. The early fault detection of the wheelset bearing is quite important to ensure the safety of the train. Among numerous diagnostic methods, envelope analysis is one of the most effective approaches in the detection of bearing faults which has been amply applied, but its validity greatly depends on the informative frequency band (IFB) determined. For the wheelset bearing faulty signal, it is often difficult to identify the IFB and extract fault characteristics due to the influence of complex operating conditions. To address this problem, a novel method to select optimal IFB, called the Mkurtogram, is proposed for railway wheelset bearings fault diagnosis. It takes the multipoint kurtosis (Mkurt) of unbiased autocorrelation (AC) of the squared envelope signal generated from sub-bands as assessment indicator for the first time. The fundamental concept which inspires this proposed method is to make full use of regular periodicity of AC of squared envelope signal. In the AC domain, the impulsiveness and periodicity, two distinctive signatures of the repetitive transients, have achieved a united representation by Mkurt. A simulated signal with multiple interferences and two experimental signals collected from wheelset bearings are applied to verify its performances and advantages. The results indicate that the proposed method is more effective to extract the wheelset bearings fault feature under complex interferences. It can not only decrease the influence of large impulse interference and the discrete harmonics interference, but also effectively overcome the influence of amplitude fluctuation caused by variable working conditions. Moreover, based on the periodic directivity of Mkurt, the proposed method also can be applied to the compound faults diagnosis of the wheelset bearing. Full article

The axle box bearing of bogie is one of the key components of the rail transit train, which can ensure the rotary motion of wheelsets and make the wheelsets adapt to the conditions of uneven railways. At the same time, the axle box bearing also exposes most of the load of the car body. Long-time high-speed rotation and heavy load make the axle box bearing prone to failure. If the bearing failure occurs, it will greatly affect the safety of the train. Therefore, it is extremely important to monitor the health status of the axle box bearing. At present, the health status of the axle box bearing is mainly monitored by vibration information and temperature information. Compared with the temperature data, the vibration data can more easily detect the early fault of the bearing, and early warning of the bearing state can avoid the occurrence of serious fault in time. Therefore, this paper is based on the vibration data of the axle box bearing to carry out adaptive fault diagnosis of bearing. First, the AR model predictive filter is used to denoise the vibration signal of the bearing, and then the signal is whitened in the frequency domain. Finally, the characteristic value of vibration data is extracted by energy operator demodulation, and the fault type is determined by comparing with the theoretical value. Through the analysis of the constructed simulation signal data, the characteristic parameters of the data can be effectively extracted. The experimental data collected from the bearing testbed of high-speed train are analyzed and verified, which further proves the effectiveness of the feature extraction method proposed in this paper. Compared with other axle box bearing fault diagnosis methods, the innovation of the proposed method is that the signal is denoised twice by using AR filter and spectrum whitening, and the adaptive extraction of fault features is realized by using energy operator. At the same time, the steps of setting parameters in the process of feature extraction are avoided in other feature extraction methods, which improves the diagnostic efficiency and is conducive to use in online monitoring system. Full article