Neural networks have been extensively applied in mechanical fault diagnosis due to their strong capabilities in feature extraction and classification. However, their limited interpretability and unknown credibility of decision hinder deployment in high-reliability scenarios. To address this issue, a frequency band multi-indicator feature embedding (FIE) method based on physical information embedding is proposed. A filter bank integrated into the convolutional layer is employed to simulate the spectrum, upon which multifrequency domain indicators are computed to extract multichannel physical features that vary continuously with frequency bands. A class-aware weight mask, enabling interclass distribution differentiation, is generated to assign distinct channel weights for different faults. This facilitates the extraction of key features for decision analysis and enables credibility evaluation based on a distance metric. Experimental results on multiple fault datasets demonstrate that the FIE module exhibits strong noise robustness and enhances diagnostic performance under variable-speed conditions. Furthermore, the class-aware mask supports reliable credibility assessment and feature interpretation, thereby supporting feature-level interpretation and decision credibility assessment while maintaining internal transparency.
XuZTangGPangB. Simulation-driven unsupervised fault diagnosis of rolling bearing under time-varying speeds. Struct Health Monit2025; 24: 3275–3291.
2.
KoTParkJJangJ, et al. Adaptive bearing fault diagnosis using reference frequency and modified scalogram. Struct Health Monit2024; 23: 2781–2794.
3.
JiaLChowTWSYuanY. Causal disentanglement domain generalization for time-series signal fault diagnosis. Neural Netw2024; 172: 106099.
4.
LuFTongQJiangX, et al. Prior knowledge embedding convolutional autoencoder: a single-source domain generalized fault diagnosis framework under small samples. Comput Ind2025; 164: 104169.
5.
HuoCXuWJiangQ, et al. An unsupervised transfer learning approach for rolling bearing fault diagnosis based on dual pseudo-label screening. Struct Health Monit2024; 23: 2288–2309.
6.
LiYMenZBaiX, et al. A bearing fault diagnosis method based on M-SSCNN and M-LR attention mechanism. Struct Health Monit2025; 24: 830–852.
7.
LiYMengXXuF, et al. Intelligent fault diagnosis of high-speed train axle box bearings under parameter fluctuation working conditions using continuous cohomology-based meta-transfer learning framework. Struct Health Monit2026; 25: 393–412.
8.
ZhaoCShenW. A domain generalization network combing invariance and specificity towards real-time intelligent fault diagnosis. Mech Syst Signal Process2022; 173: 108990.
9.
JiaLChowTWSWangY, et al. Dynamic balanced dual prototypical domain generalization for cross-machine fault diagnosis. IEEE Trans Instrum Meas2024; 73: 1–10.
10.
LeiYYangBJiangX, et al. Applications of machine learning to machine fault diagnosis: a review and roadmap. Mech Syst Signal Process2020; 138: 106587.
11.
LuFTongQJiangX, et al. Adaptive single-source open domain generalization network: a novel physical information fusion framework for fault diagnosis based on physically embedded autoencoder and adaptive open-set feature separation. Adv Eng Inform2025; 68: 103632.
12.
SiYWangYZhouD. Key-performance-indicator-related process monitoring based on improved kernel partial least squares. IEEE Trans Ind Electron2021; 68: 2626–2636.
13.
ChenH-YLeeC-H. Vibration signals analysis by explainable artificial intelligence (xai) approach: application on bearing faults diagnosis. IEEE Access2020; 8: 134246–134256.
14.
LiXChenYLiuY. A novel convolutional neural network with global perception for bearing fault diagnosis. Eng Appl Artif Intell2025; 143: 109986.
15.
QinYZhangTQianQ, et al. Large model for rotating machine fault diagnosis based on a dense connection network with depthwise separable convolution. IEEE Trans Instrum Meas2024; 73: 1–12.
16.
LiuYChenYLiX, et al. MPNet: a lightweight fault diagnosis network for rotating machinery. Measurement2025; 239: 115498.
17.
SunWYanRJinR, et al. LiteFormer: a lightweight and efficient transformer for rotating machine fault diagnosis. IEEE Trans Reliab2024; 73: 1258–1269.
18.
JiaoZPanLFanW, et al. Partly interpretable transformer through binary arborescent filter for intelligent bearing fault diagnosis. Measurement2022; 203: 111950.
19.
ShangZZhaoZYanR. Denoising fault-aware wavelet network: a signal processing informed neural network for fault diagnosis. Chin J Mech Eng2023; 36: 9.
20.
LuFTongQJiangX, et al. Wavelet packet energy embedded autoencoder with dynamic weighting strategy for fault diagnosis under unknown working conditions. Reliab Eng Syst Saf2026; 265: 111528.
21.
LiYXieSWangJ, et al. Sparse sample train axle bearing fault diagnosis: a semi-supervised model based on prior knowledge embedding. IEEE Trans Instrum Meas2023; 72: 1–11.
22.
WangZHanGLiuL, et al. A globally interpretable convolutional neural network combining bearing semantics for bearing fault diagnosis. IEEE Trans Instrum Meas2025; 74: 1–13.
23.
LiuRDingXWuQ, et al. An interpretable multiplication-convolution network for equipment intelligent edge diagnosis. IEEE Trans Syst Man Cybern Syst2024; 54: 3284–3295.
24.
HeCShiHSiJ, et al. Physics-informed interpretable wavelet weight initialization and balanced dynamic adaptive threshold for intelligent fault diagnosis of rolling bearings. J Manuf Syst2023; 70: 579–592.
25.
HeCShiHLiuX, et al. Interpretable physics-informed domain adaptation paradigm for cross-machine transfer diagnosis. Knowl Based Syst2024; 288: 111499.
26.
LuFTongQJiangX, et al. Envelope spectrum neural network with adaptive domain weight harmonization for intelligent bearing fault diagnosis under cross-machine scenarios. Adv Eng Inform2024; 62: 102787.
27.
GrezmakJZhangJWangP, et al. Interpretable convolutional neural network through layer-wise relevance propagation for machine fault diagnosis. IEEE Sens J2020; 20: 3172–3181.
28.
SanakkayalaDCVaradarajanVKumarN, et al. Explainable AI for bearing fault prognosis using deep learning techniques. Micromachines2022; 13: 1471.
29.
ShenWWeiZHuangS, et al. Interpretable compositional convolutional neural networks. Proc Int Joint Conf Artif Intell (IJCAI) 2021; 2971–2978.
30.
RavanelliMBengioY. Interpretable convolutional filters with SincNet. arXiv 2018; abs/1811.09725.
31.
ZhuQLiuHBaoC, et al. Decoupled interpretable robust domain generalization networks: a fault diagnosis approach across bearings, working conditions, and artificial-to-real scenarios. Adv Eng Inform2024; 61: 102445.
32.
LiSLiTSunC, et al. WPConvNet: an interpretable wavelet packet kernel-constrained convolutional network for noise-robust fault diagnosis. IEEE Trans Neural Netw Learn Syst2024; 35: 14974–14988.
33.
FengYZhengCChenJ, et al. Beyond deep features: fast random wavelet kernel convolution for weak-fault feature extraction of rotating machinery. Mech Syst Signal Process2025; 224: 112057.
34.
ChenQDongXTuG, et al. TFN: an interpretable neural network with time-frequency transform embedded for intelligent fault diagnosis. Mech Syst Signal Process2024; 207: 110952.
35.
DingAQinYWangB, et al. Evolvable graph neural network for system-level incremental fault diagnosis of train transmission systems. Mech Syst Signal Process2024; 210: 111175.
36.
ZhaoCZioEShenW. Domain generalization for cross-domain fault diagnosis: an application-oriented perspective and a benchmark study. Reliab Eng Syst Saf2024; 245: 109964.
37.
LiTZhaoZSunC, et al. WaveletKernelNet: an interpretable deep neural network for industrial intelligent diagnosis. IEEE Trans Syst Man Cybern Syst2022; 52: 2302–2312.
38.
FaragMM. Design and analysis of convolutional neural layers: a signal processing perspective. IEEE Access2023; 11: 27641–27661.
39.
ZhaoZJiaoY. A Fault diagnosis method for rotating machinery based on CNN with mixed information. IEEE Trans Ind Inform2023; 19: 9091–9101.
