Machines

Research

27 pages, 2172 KB

Open AccessArticle

LEADI: Operating-Mode-Aware Machine Condition Monitoring for Leak-Related Energy Anomalies—A Before-and-After Maintenance Study of a Single Production Asset

by Tanya Titova and Rosen Kosturkov

Machines 2026, 14(9), 1063; https://doi.org/10.3390/machines14091063 - 17 Sep 2026

Viewed by 61

Abstract

Compressed-air leaks create persistent parasitic demand, but machine-level condition monitoring is difficult because air consumption changes strongly with operating mode. LEADI (Leak Energy Anomaly Detection Index) was developed as an operating-mode-aware procedure that evaluates the deviation of directly measured flow rate from a [...] Read more.

Compressed-air leaks create persistent parasitic demand, but machine-level condition monitoring is difficult because air consumption changes strongly with operating mode. LEADI (Leak Energy Anomaly Detection Index) was developed as an operating-mode-aware procedure that evaluates the deviation of directly measured flow rate from a local reference baseline derived from a stable post-repair condition with maintained pressure and low within-window variability. The method was developed on days 1–5 and evaluated on held-out days 6–7 from two one-week campaigns conducted before and after implementation of the prescribed corrective actions. With 60 min windows, LEADI flagged 19/19 evaluable pre-repair and 0/17 post-repair windows, with diagnostic coverage of 39.6% and 35.4%, respectively. A simple fifth-percentile flow comparator without operating-mode selection flagged 47/48 versus 1/48 windows. This shows that the low-flow region itself contains strong discriminatory information for separating the two periods. The role of the operating-mode layer is to restrict engineering interpretation to pre-specified eligible operating conditions. The flow-rate difference within the diagnostic operating condition was 122.0 L/min (95% CI 114.6–132.3). Over a common 168 h basis, measured volume decreased by 1370.8 m³ (28.90%), while a separate check normalized by pressurized time gave 28.12%. Because the specific energy consumption of the compressor station was not measured, the energy effect is reported only as a scenario for the same 168 h. The results support the applicability of LEADI as a selective decision-support layer for the investigated asset and the two observed conditions, without establishing universal leak detection or causal attribution of the observed change to individual defects. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

33 pages, 12143 KB

Open AccessArticle

Ensemble Network-State Forecasting for Remote Fault Diagnosis Using Transformer and Ridge Regression

by Zehua Sun, Yancai Xiao, Haikuo Shen and Shaodan Zhi

Machines 2026, 14(9), 1037; https://doi.org/10.3390/machines14091037 - 11 Sep 2026

Viewed by 143

Abstract

Remote fault-diagnosis services in dynamic edge–cloud environments depend on timely monitoring-data upload, remote inference, and result delivery, making their communication layer sensitive to variations in available bandwidth, link latency, and packet loss rate. This study addresses the network-state forecasting layer that supports such [...] Read more.

Remote fault-diagnosis services in dynamic edge–cloud environments depend on timely monitoring-data upload, remote inference, and result delivery, making their communication layer sensitive to variations in available bandwidth, link latency, and packet loss rate. This study addresses the network-state forecasting layer that supports such services rather than the fault-classification model itself. We propose Horizon-Aware Transformer–Ridge Fusion (HATR-Fusion), which combines a nonlinear Transformer expert with a low-variance ridge-regression expert for joint short- and long-horizon forecasting. Historical available bandwidth, link latency, packet loss rate, mobility, and offered load are used as inputs. Preprocessing statistics are estimated using the training set only, and validation-calibrated convex fusion weights are frozen before test inference. Experiments on controlled synthetic trajectories from eight links sampled at 1-min intervals, using 10 neural-network initialization seeds and ten baselines including DLinear and iTransformer, show that HATR-Fusion reduces mean absolute error (MAE) relative to the standalone Transformer by 6.94–8.31% over the 10-min horizon and by 3.02–4.40% over the 60-min horizon, with all six paired improvements remaining significant after Holm correction. Against iTransformer, HATR-Fusion is significantly more accurate for short-horizon bandwidth and latency, whereas iTransformer is significantly more accurate for long-horizon latency and packet loss; short-horizon packet loss and long-horizon bandwidth are not significantly different after Holm correction. The six-task mean normalized mean absolute error (NMAE) is 0.07106 for HATR-Fusion and 0.07047 for iTransformer, indicating comparable overall accuracy with task-dependent differences between the two methods. Ablation results show complementary short- and long-range contributions from ridge regression and Transformer, while input-quality sensitivity analysis identifies a limitation of the fixed fusion weights under corrupted or missing history. The conclusions are therefore restricted to scenarios with relatively stable input quality and distribution shifts comparable to those evaluated in this study. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

22 pages, 4971 KB

Open AccessArticle

Temperature Field-Based Detection of Oil Supply Failure in Plain Bearings Considering Varying Component Sizes

by Thao Baszenski, Karl-Heinz Kratz, Georg Jacobs, Tobias Gemmeke, Benjamin Lehmann and Mattheüs Lucassen

Machines 2026, 14(9), 962; https://doi.org/10.3390/machines14090962 - 25 Aug 2026

Viewed by 307

Abstract

Plain bearings are widely used in heavy-duty applications, e.g., wind turbine drivetrains or ship propulsion systems. Plain bearings offer high load-carrying capacity and good damping, but abnormal events such as oil supply failure (OSF) can rapidly damage the bearing and cause failure of [...] Read more.

Plain bearings are widely used in heavy-duty applications, e.g., wind turbine drivetrains or ship propulsion systems. Plain bearings offer high load-carrying capacity and good damping, but abnormal events such as oil supply failure (OSF) can rapidly damage the bearing and cause failure of the entire drivetrain. An adequate oil supply is essential for the operation of the plain bearing. A failure of the oil supply can cause fatal failure of the bearing due to adhesive wear within a matter of seconds. Existing condition monitoring systems (CMS) for plain bearings generally cannot detect OSF in time, or require costly and complex installation, limiting their practical applicability. This paper presents temperature field measurement (TFM) as a simple, low-cost CMS approach for early OSF detection. The results presented within this work demonstrate that TFM detects the onset of OSF within 30 s of oil supply interruption, at least 30 s before a rise in friction torque can be detected in the corresponding test bearing. The findings demonstrate that TFM is a valid, simple, and effective method for the timely detection of OSF, offering a practical alternative to existing CMS approaches for heavy-duty plain bearing applications. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

26 pages, 4996 KB

Open AccessArticle

Protocol and Implementation Sensitivity in Raw-Index-Audited Few-Shot Bearing Fault Diagnosis Benchmarking: Evidence from CWRU and HUSTbearing

by Jianxin Zhang and Guixiang Shen

Machines 2026, 14(8), 891; https://doi.org/10.3390/machines14080891 - 5 Aug 2026

Viewed by 395

Abstract

Sliding-window few-shot evaluations can place support and query windows over shared raw samples. We audited episode-internal overlap, source-noise policy, normalization, and fixed-feature construction using Case Western Reserve University (CWRU) data and a HUSTbearing cross-speed task reconstructed from raw files. In the primary CWRU [...] Read more.

Sliding-window few-shot evaluations can place support and query windows over shared raw samples. We audited episode-internal overlap, source-noise policy, normalization, and fixed-feature construction using Case Western Reserve University (CWRU) data and a HUSTbearing cross-speed task reconstructed from raw files. In the primary CWRU 0 hp to 3 hp (0 to approximately 2.24 kW), 10-way, 5-shot evaluation at a signal-to-noise ratio (SNR) of −5 dB across 20 seeds, common-pool construction increased log-compressed fast Fourier transform prototype (Log-FFT) accuracy by 3.83 percentage points (95% confidence interval (CI): [+3.63, +4.02]). The corresponding increase for the clean source-supervised cross-entropy prototype encoder (Source-CE-Proto) was 0.95 percentage points and was not statistically distinguishable from zero (95% CI: [−0.004, +1.90]; exact p = 0.05084). Under raw-index-separated evaluation, leave-one-SNR-out Source-CE-Proto achieved 93.23% without direct −5 dB source exposure, compared with 82.43% for Log-FFT. A No-log FFT control achieved 99.09% in the primary task and also exceeded the learned encoders in a second CWRU load pair. HUSTbearing likewise showed channel-dependent fixed-feature accuracy and overlap effects. These results do not support a general method ranking; they support reporting raw-index provenance, source-noise distributions, normalization, and exact feature construction before interpreting method rankings in few-shot bearing fault diagnosis benchmarks. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

26 pages, 6084 KB

Open AccessArticle

A PINN-Based Fault Diagnosis Method for Crack Damage in Wind Turbine Blades

by Min Wang, Guo-Jun Qin and Xiao-Fei Zhang

Machines 2026, 14(8), 857; https://doi.org/10.3390/machines14080857 - 28 Jul 2026

Viewed by 561

Abstract

Vibration response analysis constitutes a pivotal approach for crack monitoring and early warning of damage identification in wind turbine blades. Traditional data-driven methods, however, demonstrate marked deficiencies in identification accuracy and generalization capability. To mitigate these issues, a method for crack damage identification [...] Read more.

Vibration response analysis constitutes a pivotal approach for crack monitoring and early warning of damage identification in wind turbine blades. Traditional data-driven methods, however, demonstrate marked deficiencies in identification accuracy and generalization capability. To mitigate these issues, a method for crack damage identification in wind turbine blades is proposed, grounded in Physics-Informed Neural Networks (PINNs). Initially, utilizing a scaled-down test platform for doubly fed wind turbines, simulation experiments on blade cracks were executed. Vibration data were amassed under varying crack locations and lengths to scrutinize the intrinsic relationship between crack characteristics and the three-dimensional vibration response of the blade root bearing pedestal. Subsequently, leveraging the rotating cantilever Euler–Bernoulli beam model, the physical correlation between cracks and vibrations was dissected, and a physical information constraint model was formulated. This model was then amalgamated with a GRU-Transformer network to establish a PINN model tailored for crack damage identification. Ultimately, the model underwent testing and validation utilizing experimental data. The outcomes reveal that, in comparison to traditional data-driven models, the PINN model exhibits superior accuracy and precision in crack identification and localization, along with exceptional generalization capability and noise resilience. This research provides a novel technical pathway for enhancing the intelligence level of health monitoring for wind turbine units and holds substantial engineering significance for achieving precise condition assessment and early fault warning. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

24 pages, 1404 KB

Open AccessArticle

An Acoustic Fault Diagnosis Method for Oil and Gas Pipelines Based on Time–Frequency Diagrams and Parallel CNN-GRU

by Yang Peng, Shaomu Wen, Yongbo Wang, Kedu Ma, Qin Bie and Wei He

Machines 2026, 14(8), 846; https://doi.org/10.3390/machines14080846 - 27 Jul 2026

Viewed by 420

Abstract

Oil and gas pipelines are the core infrastructure of energy transportation, and their safe operation is crucial to national energy security. Aiming at the difficulty of feature extraction and insufficient diagnosis accuracy of pipeline acoustic fault, a fault diagnosis method based on dual-branch [...] Read more.

Oil and gas pipelines are the core infrastructure of energy transportation, and their safe operation is crucial to national energy security. Aiming at the difficulty of feature extraction and insufficient diagnosis accuracy of pipeline acoustic fault, a fault diagnosis method based on dual-branch parallel feature fusion of the original time-series signal and time–frequency map was proposed. In this method, the time–frequency map of the one-dimensional acoustic signal was generated by continuous wavelet Transform (CWT), and the original signal was input into the dual-branch network, respectively. The spatial–frequency domain features were extracted by using lightweight depthwise separable convolution (LDconv) embedded with coordinate attention (CA) in the upper branch. The lower branch mines local details and temporal dependencies through deformable convolution v4 (DCNv4) and Gated Recurrent Unit (GRU). The dual-branch features were concatenated and fused by Global Average Pooling (GAP), and finally the classification results were output by the fully connected network and Softmax. Experiments on industrial field data show that the average diagnostic accuracy of the proposed method is 98.87%, which can effectively extract weak fault features under complex noise, and has significant advantages in early fault recognition and generalization performance. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

23 pages, 27297 KB

Open AccessArticle

CWT-PSDT-Based Identification of Electromagnetic-Related Stator Vibration Frequency Components in a Hydro-Generator

by Jiannan Zhao, Juan Duan, Kun Yang, Jianlan Wang, Junqing Wang, Xuan Yang and Jiacai Feng

Machines 2026, 14(7), 807; https://doi.org/10.3390/machines14070807 - 16 Jul 2026

Viewed by 397

Abstract

Accurate identification of electromagnetically induced stator vibration frequency components is essential for the online condition monitoring of hydro-generators, particularly for assessing the dynamic state of the stator core under normal operating conditions. In engineering practice, the fast Fourier transform (FFT) is widely used [...] Read more.

Accurate identification of electromagnetically induced stator vibration frequency components is essential for the online condition monitoring of hydro-generators, particularly for assessing the dynamic state of the stator core under normal operating conditions. In engineering practice, the fast Fourier transform (FFT) is widely used for vibration spectrum analysis; however, because the measured vibration response is simultaneously affected by electromagnetic excitation, mechanical rotation, hydraulic disturbance, and external harmonic interference, FFT-based spectra often contain multiple frequency components whose structural relevance is difficult to determine directly. To address this issue, this paper proposes a coupled continuous wavelet transform and power spectral density transmissibility (CWT-PSDT) method for identifying key vibration frequency components with stable time-frequency energy and inter-sensor transmissibility in hydro-generator stator vibration signals. In the proposed framework, the analytic Morlet wavelet is first employed to localize dominant energy bands in the time-frequency domain, and PSDT is then used to screen frequency components with relatively stable inter-sensor transmissibility characteristics, thereby reducing the ambiguity caused by excitation-dominated spectral components. A clamped-clamped beam model is first used for numerical validation, and the maximum identification error of the first five natural frequencies is 4.22%. Experiments on a Francis turbine-generator test rig under five operating conditions further show that the proposed method can distinguish the mechanical rotational component near 10.3 Hz from the electromagnetic-related component near 50.8 Hz, while retaining higher-order electromagnetic-related components around 150 Hz and 250 Hz. The results demonstrate that the proposed CWT-PSDT method provides a physically interpretable and data-efficient approach for extracting stator-core-related spectral features, and offers a theoretical basis for spectrum-based online monitoring and future abnormal-condition comparison of hydro-generator stator responses. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

21 pages, 4536 KB

Open AccessArticle

Partial Discharge Severity Classification for Transformer Condition Monitoring Using Feature Engineering, PCA, and ANN

by Lucas Thobejane and Bonginkosi A. Thango

Machines 2026, 14(6), 711; https://doi.org/10.3390/machines14060711 - 22 Jun 2026

Viewed by 472

Abstract

Partial discharge (PD) is a key indicator of insulation degradation in high-voltage transformers and can provide early warning of incipient failure. Although artificial neural networks (ANNs) have been applied to PD classification, their performance may be affected by redundant features and overfitting when [...] Read more.

Partial discharge (PD) is a key indicator of insulation degradation in high-voltage transformers and can provide early warning of incipient failure. Although artificial neural networks (ANNs) have been applied to PD classification, their performance may be affected by redundant features and overfitting when using expanded feature spaces. This study proposes a PD severity classification framework that combines physics-informed feature engineering, principal component analysis (PCA), and a multilayer perceptron (MLP) neural network. PD measurements were acquired from a physical transformer using the IEC 60270 electrical measurement method, yielding 294 samples labelled into four severity classes: normal, low, medium, and high PD. Two measured variables, namely PD magnitude and applied voltage, were expanded into a 10-dimensional feature space using energy-based, ratio-based, logarithmic, and normalized features. PCA was then used to reduce the feature space, and the retained principal components were used as inputs to the classifier. The results show that the first two principal components captured more than 90% of the total variance and enabled the MLP to achieve 98.3% test accuracy, matching the performance obtained using all 10 engineered features and improving on classification based on the raw measurements alone (91.5%). The proposed PCA-ANN model also achieved perfect precision and recall for the medium- and high-severity classes on the test set, and outperformed K-nearest neighbours, support vector machine, and Gaussian Naïve Bayes models in 5-fold cross-validation. These findings indicate that PCA can reduce feature dimensionality without loss of diagnostic performance, providing an efficient approach for transformer PD severity classification. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

20 pages, 4211 KB

Open AccessArticle

On the Role of Feature Extraction in Transformer PD Severity Classification: A Controlled Comparison of PCA and Autoencoder Models

by Lucas Thobejane and Bonginkosi Thango

Machines 2026, 14(6), 708; https://doi.org/10.3390/machines14060708 - 21 Jun 2026

Cited by 1 | Viewed by 425

Abstract

This paper applies the comparative PCA-ANN vs. Autoencoder-ANN framework to transformer partial discharge (PD) severity classification, using a 294-sample dataset spanning four severity classes: Normal, Low PD, Medium PD, and High PD. Two raw measurements of discharge magnitude (pC) and applied voltage (kV) [...] Read more.

This paper applies the comparative PCA-ANN vs. Autoencoder-ANN framework to transformer partial discharge (PD) severity classification, using a 294-sample dataset spanning four severity classes: Normal, Low PD, Medium PD, and High PD. Two raw measurements of discharge magnitude (pC) and applied voltage (kV) are expanded into a 15-dimensional physics-informed feature space. Both linear (PCA) and nonlinear (bottleneck Autoencoder) feature extraction are evaluated exhaustively across all latent dimensions k = 1–15, feeding an identical ANN classifier. PCA + ANN achieves perfect test accuracy of 100.0% at k = 9, while Autoencoder + ANN achieves 98.3% at k = 8. PCA + ANN demonstrates superior performance on this dataset, attributed to the low intrinsic dimensionality of the two-measurement PD feature space and the highly separable nature of PD severity classes in the engineered ratio feature space. The Autoencoder provides a more compact latent representation but introduces classification errors for the Normal class due to its extreme under-representation. Cross-validation confirms PCA + ANN stability (97.4 ± 0.9% vs. 97.0 ± 1.0%). These results, alongside the companion DGA study, provide the complete baseline for comparing linear and nonlinear feature extraction across two transformer diagnostic modalities. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

36 pages, 2937 KB

Open AccessArticle

BIM and PdM of Railway Rolling Stock with Automatic Upgrading Based on GenAI

by João Matos Coutinho, Hugo Raposo, José M. Torres Farinha and Antonio J. Marques Cardoso

Machines 2026, 14(5), 535; https://doi.org/10.3390/machines14050535 - 11 May 2026

Viewed by 1923

Abstract

The paradigm transition of the life cycle management of physical assets in the railway sector demands new maintenance models that imply the conventional predictive approaches to be surpassed. This paper proposes an innovative methodology that integrates Building Information Modelling (BIM) with predictive maintenance [...] Read more.

The paradigm transition of the life cycle management of physical assets in the railway sector demands new maintenance models that imply the conventional predictive approaches to be surpassed. This paper proposes an innovative methodology that integrates Building Information Modelling (BIM) with predictive maintenance (PdM) systems to be applied to rolling stock and, in this way, be enhanced by Generative Artificial Intelligence (GenAI). The research focuses on the autonomous synchronisation of the Rolling Stock Digital Twin (DT). Unlike static BIM models, the proposed solution enables the use of GenAI algorithms to process continuous data streams from integrated sensors, allowing the digital model to evolve autonomously as physical wear occurs. In this framework, GenAI (via Generative Adversarial Networks—GANs) is essential for data augmentation, enabling the simulation of rare “long-tail” failure events that are scarce in real-world historical data. By synthesising these degradation scenarios, the model learns complex mechanical collapse patterns that otherwise would be ignored by traditional PdM approaches. GenAI is employed to synthesise degradation scenarios, perform real-time parametric updates within the IFC (Industry Foundation Classes) schema, and optimise maintenance workflows. The application of this framework demonstrates a significant reduction in diagnostic latency and optimises the rolling stock’s operational life cycle by automating updates and reducing the need for manual data entry. This study concludes that the convergence among BIM, PdM, and GenAI establishes a robust framework for railway fleet management. While the current validation focuses on bogie systems using Random Forest and LLMs, it paves the way for a future Industrial Metaverse where immersive diagnostics can be integrated into the maintenance lifecycle. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

23 pages, 5737 KB

Open AccessArticle

Efficient Dual-Stream Network with Soft-Gated Fusion for Bearing Fault Diagnosis Using Acoustic Emission Signals

by Van-Loc Le, Huynh-Anh-Huy Nguyen and Cheol Hong Kim

Machines 2026, 14(4), 414; https://doi.org/10.3390/machines14040414 - 8 Apr 2026

Viewed by 1163

Abstract

Bearings play crucial roles in industrial machinery. Therefore, the continuous monitoring and effective detection of bearing failures are essential to ensure the safety and reliability of motors. Traditional fault diagnosis methods often require information from both the time and frequency domains; however, converting [...] Read more.

Bearings play crucial roles in industrial machinery. Therefore, the continuous monitoring and effective detection of bearing failures are essential to ensure the safety and reliability of motors. Traditional fault diagnosis methods often require information from both the time and frequency domains; however, converting them into a two-dimensional representation significantly increases computational costs. Conversely, utilizing only time-domain features while ignoring frequency-domain features results in incomplete fault information, reducing accuracy under various operating conditions. This study proposes an efficient dual-stream network with soft-gated fusion for bearing fault diagnosis that simultaneously analyzes acoustic emission signals in the time and frequency domains. Our approach employs two separate feature-learning branches: the time-domain branch directly extracts features from the segmented raw acoustic emission signals, and the frequency-domain branch learns features from one-dimensional spectral vectors obtained using the fast Fourier transform. A gated fusion mechanism adaptively balances the contribution of each domain before classifying fault types. The experimental results show that the proposed method significantly reduces the computational cost compared with that of a two-dimensional-representation-based model and improves accuracy over time-only or frequency-only baselines. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

21 pages, 2216 KB

Open AccessArticle

Lightweight MS-DSCNN-AttMPLSTM for High-Precision Misalignment Fault Diagnosis of Wind Turbines

by Xiangyang Zheng, Yancai Xiao and Xinran Li

Machines 2026, 14(2), 155; https://doi.org/10.3390/machines14020155 - 29 Jan 2026

Cited by 1 | Viewed by 870

Abstract

Wind turbine (WT) misalignment fault diagnosis is constrained by critical signal processing challenges: weak fault features, intense background noise, and poor generalization. This study proposes a lightweight method for high-precision fault diagnosis. A fixed-threshold wavelet denoising method with the scene-specific pre-optimized parameter a [...] Read more.

Wind turbine (WT) misalignment fault diagnosis is constrained by critical signal processing challenges: weak fault features, intense background noise, and poor generalization. This study proposes a lightweight method for high-precision fault diagnosis. A fixed-threshold wavelet denoising method with the scene-specific pre-optimized parameter a (0 < a ≤ 1.3) is proposed: the parameter a is determined via offline grid search using the feature retention rate (FRR) as the objective function for typical wind farm operating scenarios. A multi-scale depthwise separable CNN (MS-DSCNN) captures multi-scale spatial features via 3 × 1 and 5 × 1 kernels, reducing computational complexity by 73.4% versus standard CNNs. An attention-based minimal peephole LSTM (AttMPLSTM) enhances temporal feature measurement, using minimal peephole connections for long-term dependencies and channel attention to weight fault-relevant signals. Joint L1–L2 regularization mitigates overfitting and environmental interference, improving model robustness. Validated on a WT test bench, the Adams simulation dataset, and the CWRU benchmark, the model achieves a 90.2 ± 1.4% feature retention rate (FRR) in signal processing, an over 98% F1-score for fault classification, and over 99% accuracy. With 2.5 s single-epoch training and a 12.8 ± 0.5 ms single-sample inference time, the reduced parameters enable real-time deployment in embedded systems, advancing signal processing for rotating machinery fault diagnosis. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

21 pages, 1371 KB

Open AccessArticle

Speed Independent Health Indicator for Outer Raceway Bearing Fault Using MCSA

by Praneet Amitabh, Dimitar Bozalakov and Frederik De Belie

Machines 2025, 13(12), 1095; https://doi.org/10.3390/machines13121095 - 26 Nov 2025

Viewed by 971

Abstract

Bearing health monitoring is essential for ensuring the reliability and operational safety of induction machines, as bearing faults remain among the most frequent failure modes in rotating electrical equipment. This work contributes to condition monitoring by enhancing the robustness of health indicators and [...] Read more.

Bearing health monitoring is essential for ensuring the reliability and operational safety of induction machines, as bearing faults remain among the most frequent failure modes in rotating electrical equipment. This work contributes to condition monitoring by enhancing the robustness of health indicators and developing a supply-frequency-independent health indicator (HI) for bearing fault diagnosis using Motor Current Signature Analysis (MCSA). The objective is to design an HI capable of reliably representing the bearing degradation state under varying operating conditions, particularly when the supply frequency changes. To achieve this, the study briefly examines the key physical mechanisms governing the detectability of bearing-related spectral signatures—including rotational frequency, unbalanced magnetic pull, eddy currents, skin effect, and hydrodynamic forces. The theoretical analysis establishes the overall trend expected under varying supply frequencies and clarifies how these phenomena collectively influence the spectral characteristics of the fault components and the frequency-dependent evolution of their amplitudes. These insights are experimentally validated using induction machines fitted with bearings of two fault severities. Leveraging this physical understanding, a modified regression-based compensation model is introduced to reduce the frequency-dependent variation in the HI. The resulting compensating factor effectively stabilizes the frequency response, producing a more consistent and monotonic degradation trend across the tested conditions. The proposed method is computationally lightweight, does not require run-to-failure data or detailed physical modeling, and is suitable for real-time implementation. By integrating physical insight with data-driven modeling, this work presents a practical and frequency-independent HI framework that can be readily deployed within digital-twin-based condition monitoring architectures for induction machines. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

16 pages, 6172 KB

Open AccessArticle

A Novel Dual-Channel Hybrid Attention Model for Wind Turbine Misalignment Fault Diagnosis

by Tong Tong, Xiang Liu, Jia Zhang, Dian Long, Teng Fan and Xiangyang Zheng

Machines 2025, 13(5), 368; https://doi.org/10.3390/machines13050368 - 29 Apr 2025

Cited by 1 | Viewed by 977

Abstract

Aiming at the problems of inaccurate feature extraction, slow convergence, and low diagnostic accuracy of wind turbine misalignment fault diagnosis under complex working conditions, this paper proposes an innovative diagnostic method based on two channels of U-Net and ResNet50. The model innovatively introduces [...] Read more.

Aiming at the problems of inaccurate feature extraction, slow convergence, and low diagnostic accuracy of wind turbine misalignment fault diagnosis under complex working conditions, this paper proposes an innovative diagnostic method based on two channels of U-Net and ResNet50. The model innovatively introduces the multi-head attention mechanism (MHA) in the jump connection of the U-Net architecture to form hybrid U-Net and optimizes the feature fusion process with dynamically learnable weights, which significantly enhances the ability to capture local details and key fault features. In the ResNet50 branch, deep global features are fully mined for extraction. To further achieve the co-optimization of global and local information, a shared hybrid expert attention (SHEA) module is proposed. This module achieves efficient integration of features by adaptively fusing the multi-scale local features output from the hybrid U-Net decoder with the deep global features extracted from the ResNet50 backbone network through a dynamic weighting and expert selection mechanism. The multi-scale features optimized by the SHEA module are fed into the classifier for fault type determination. The experimental results show that the method demonstrates excellent convergence speed and 99.64% classification accuracy under complex working conditions, providing an effective solution for the intelligent diagnosis of wind turbine misalignment faults. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

32 pages, 15651 KB

Open AccessArticle

Intelligent Diagnosis of Rolling Element Bearings Under Various Operating Conditions Using an Enhanced Envelope Technique and Transfer Learning

by Ali Davoodabadi, Mehdi Behzad, Hesam Addin Arghand, Somaye Mohammadi and Len Gelman

Machines 2025, 13(5), 351; https://doi.org/10.3390/machines13050351 - 23 Apr 2025

Cited by 5 | Viewed by 1783

Abstract

Rolling element bearings (REBs) are vital in rotating machinery, making fault detection essential for optimal performance and system reliability. This study assesses the effectiveness of a simple convolutional neural network (SCNN) and a transfer learning-based convolutional neural network (TL-CNN) for diagnosing REB faults [...] Read more.

Rolling element bearings (REBs) are vital in rotating machinery, making fault detection essential for optimal performance and system reliability. This study assesses the effectiveness of a simple convolutional neural network (SCNN) and a transfer learning-based convolutional neural network (TL-CNN) for diagnosing REB faults using time-domain signals, frequency-domain spectra, and envelope frequency spectrum analysis. The study uses diverse datasets, including laboratory and industrial data under various operating conditions, covering fault types like inner race fault (IRF), outer race fault (ORF), rolling element fault (REF), and healthy (H) states. The main innovation is applying Transfer Learning (TL) with fine-tuning to improve model accuracy in identifying REB conditions by leveraging features learned from diverse datasets. An innovative algorithm is also introduced to identify resonance regions for optimal filter selection in envelope analysis, improving fault-related feature extraction and reducing noise. A preprocessing step that removes speed-related variations further enhances model accuracy by isolating fault features and minimizing the impact of rotational speed. The results show that transfer learning with fine-tuning, combined with the resonance region identification algorithm, significantly enhances fault detection accuracy. The TL-CNN model with envelope signal input achieves the highest accuracy across all scenarios, especially under variable operating conditions, and performs reliably on industrial data. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Graphical abstract

15 pages, 5515 KB

Open AccessArticle

Combining Sensor Fusion and a Machine Learning Framework for Accurate Tool Wear Prediction During Machining

by Swathi Kotha Amarnath, Vamsi Inturi, Sabareesh Geetha Rajasekharan and Amrita Priyadarshini

Machines 2025, 13(2), 132; https://doi.org/10.3390/machines13020132 - 10 Feb 2025

Cited by 20 | Viewed by 4633

Abstract

Effective cutting tool condition monitoring (TCM) is critical for achieving precision, cost efficiency, and minimizing unplanned downtime. This study proposes a sophisticated sensor fusion framework for accurate tool fault prediction during machining. Experimental data were collected while turning AISI 410-grade steel bars with [...] Read more.

Effective cutting tool condition monitoring (TCM) is critical for achieving precision, cost efficiency, and minimizing unplanned downtime. This study proposes a sophisticated sensor fusion framework for accurate tool fault prediction during machining. Experimental data were collected while turning AISI 410-grade steel bars with uncoated carbide inserts under dry-cutting conditions. Force and vibration signals were captured across five tool health states (one healthy and four faulty) using a sensor network and data acquisition systems. The raw signals were decomposed using discrete wavelet transform, and key statistical features were extracted. Three distinct input datasets are constructed: Dataset I comprises statistical parameters extracted exclusively from the force signals, Dataset II consists of statistical parameters derived from the vibration signals, and Dataset III integrates the individual statistical parameters from both force and vibration signals through feature-level fusion. These datasets are then utilized for training ML classifiers (Support Vector Machine, Random Forest, and Naive Bayes) to perform feature learning and subsequent classification. Among the considered classifiers, the RF classifier yielded better classification accuracies of 96% and 97% while discriminating among the tool health scenarios through dataset I and II. Also, the RF and SVM classifiers achieved a classification accuracy of 98% and 88% in distinguishing tool health scenarios for dataset III. This method demonstrates exceptional suitability for real-time, in situ fault diagnostics and provides a strong foundation for developing online TCM systems, advancing the objectives of Industry 4.0 and smart manufacturing. Full article

(This article belongs to the Special Issue Condition Monitoring and Fault Diagnosis)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Condition Monitoring and Fault Diagnosis

Share This Special Issue

Editors

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (16 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI