| 93 | 0 | 26 |
| 下载次数 | 被引频次 | 阅读次数 |
针对铁路轴承在强噪声与复杂工况下诊断精度不足的问题，提出了一种融合马尔可夫转移场（MTF）、格拉姆角场（GAF）与短时傅里叶变换（STFT）的多域特征并行提取方法。通过小波分解实现振动信号降噪，分别在时域、角域与频域构建特征图像，经卷积神经网络提取深层特征后进行融合，并通过麻雀搜索算法（SSA）调节XGBoost模型超参数，以增强分类器的泛化与稳定性。通过CWRU轴承数据集验证，该方法在多类故障识别中的准确率达到99.81%，并在低信噪比条件下仍可保持较高鲁棒性，为铁路装备的状态监测和早期故障预警提供了一种可行的技术方案。
Abstract：To address the challenges of low diagnostic accuracy and feature loss in railway bearing fault diagnosis under strong noise and complex operating conditions, a multi-domain feature fusion method combining Markov Transition Field(MTF),Gramian Angular Field(GAF),and Short-Time Fourier Transform(STFT) is proposed.Wavelet decomposition is first applied for signal denoising to retain critical transient features. The denoised signals are then mapped into time,angular,and frequency domains,from which deep features are extracted via convolutional neural networks(CNNs) and fused for comprehensive representation. In the classification stage,the Sparrow Search Algorithm(SSA) is employed to globally optimize the hyperparameters of the XGBoost classifier, enhancing its generalization and robustness. Experiments on the CWRU bearing dataset demonstrate that the proposed method achieves an accuracy of 99.81% and maintains strong robustness under low signal-to-noise conditions(–6 dB). This framework provides a practical and intelligent solution for the condition monitoring and early fault diagnosis of railway bearings.
[1]Xiao L, Yang X, Yang X. A graph neural network-based bearing fault detection method[J]. Scientific Reports, 2023, 13(1):5286.
[2]乔志城，刘永强，廖英英.改进经验小波变换与最小熵解卷积在铁路轴承故障诊断中的应用[J].振动与冲击，2021,40(2):81-90.QIAO Zhicheng,LIU Yongqiang,LIAO Yingying. Application of improved empirical wavelet transform and minimum entropy deconvolution in railway bearing fault diagnosis[J]. Vibration and Shock,2021,40(2):81-90.
[3]赵恒喆，杨晓英，汤洁，等.高速铁路动车组轴箱轴承故障分析[J].轴承，2022(7):1-8.ZHAO Hengzhe,YANG Xiaoying,TANG Jie,et al. Analysis of axle-box bearing faults of high-speed trainsets[J]. Bearings,2022(7):1-8.
[4]邓飞跃，吕浩洋，顾晓辉，等.基于轻量化神经网络Shuffle-SENet的高速动车组轴箱轴承故障诊断方法[J].吉林大学学报(工学版)，2022,52(2):474-482.DENG Feiyue,LYU Haoyang,GU Xiaohui,et al. Bearing fault diagnosis method for high-speed EMU axle-box based on lightweight neural network Shuffle-SENet[J]. Journal of Jilin University(Engineering and Technology Edition),2022,52(2):474-482.
[5]Cai C,Xu T,Ren J,et al. Bearing fault diagnosis based on the Markov transition field and SE-IShufflenetV2 model[J]. Structural Durability&Health Monitoring,2025,19(1):21-39.
[6]Feng R,Zhang Q,Wang L,et al. Rolling bearing fault diagnosis method based on MTF-MMCNN[J]. Nondestructive Testing and Evaluation,2025:1-21.
[7]Yan J,Kan J,Luo H,et al. Rolling bearing fault diagnosis based on Markov transition field and residual network[J]. Sensors,2022,22(10):3936.
[8]Shen J,Wu Z,Cao Y,et al. Research on fault diagnosis of rolling bearing based on gramian angular field and lightweight model[J].Sensors,2024,24(18):5952.
[9]Liu C,Bai J,Xue L,et al. Rolling bearing fault diagnosis method based on gramian angular difference field and dynamic self-calibrated convolution module[J]. PloS one,2024,19(12):e0314898.
[10]Han B,Zhang H,Sun M,et al. A new bearing fault diagnosis method based on capsule network and Markov transition field/Gramian angular field[J]. Sensors,2021,21(22):7762.
[11]Xiang C,Zhou J,Han B,et al. Fault diagnosis of rolling bearing based on a priority elimination method[J]. Sensors,2023,23(4):2320.
[12]Zhong R,Li C,Wang M,et al. Fault diagnosis of rolling bearings based on ISSA-XGBoost with FLDA feature dimensionality reduction[C]//2025 International Conference on Mechatronics,Robotics,and Artificial Intelligence(MRAI). IEEE,2025:437-446
[13]Shi W,Chen L. Wheat freshness recognition leveraging Gramian angular field and attention-augmented ResNet[J]. Open Agriculture,2025,10(1):20250437.
[14]张又文，冯斌，陈页，等.基于遗传算法优化XGBoost的油浸式变压器故障诊断方法[J].电力自动化设备，2021,41(2):7-13.ZHANG Youwen,FENG Bin,CHEN Ye,et al. Transformer fault diagnosis for oil-immersed transformers based on GA-optimized XGBoost[J]. Power Automation Equipment,2021,41(2):7-13.
[15]贾皓阳，钱宇.基于贝叶斯优化XGBoost算法的变压器故障诊断[J].黄河水利职业技术学院学报，2023,35(2):37-43.JIA Haoyang , QIAN Yu. Transformer fault diagnosis based on bayesian-optimized XGBoost[J]. Journal of Yellow River Conservancy Technical Institute,2023,35(2):37-43.
[16]段震清，孙建民，梁凌，等.基于XGBoost算法的光伏阵列故障诊断方法研究[J].电力系统自动化，2023,47(4):112-119.DUAN Zhenqing , SUN Jianmin , LIANG Ling ,et al. Research on PV array fault diagnosis method based on XGBoost[J].Automation of Electric Power Systems,2023,47(4):112-119.
[17]刘静，李兵.滚动轴承内部缺陷动力学建模与数值仿真[M].西北工业大学出版社，2022.
[18]Kim S H,Jung W,Lim D,et al. Fault diagnosis of ball bearing using dynamic convolutional neural networks under varying speed condition[C]//Proc. IECON 2022–48th Annual Conference of the IEEE Industrial Electronics Society. IEEE,2022:1-6.
[19]Ding H,Cheng X,Wang K,et al. Fault feature extraction of rolling bearing based on maximum second-order cyclostationarity blind deconvolution and CEEMD-teager[J]. International Journal of Acoustics&Vibration,2024,29(3):235.
[20]Pei D,Yue J,Jiao J. A novel method for bearing fault diagnosis under variable speed based on envelope spectrum fault characteristic frequency band identification[J]. Sensors,2023,23(9):4338.
[21]Smith W A,Randall R B. Rolling element bearing diagnostics using the Case Western Reserve University data:a benchmark study[J].Mechanical Systems and Signal Processing,2015,64:100-131.
基本信息:
DOI：10.20213/j.cnki.tdcl.2026.09.29.01
中图分类号:U279
引用信息:
[1]谭浩宇,刘武军,刘思源,等.基于MTF-GAF-STFT的时频特征融合轴承故障诊断方法[J].铁道车辆().DOI:10.20213/j.cnki.tdcl.2026.09.29.01.
基金信息:
湖南省自然科学基金资助项目（2025JJ80335）; 湖南省教育厅科学研究项目（23B1020）
2026-09-30
2026-09-30
2026-09-30