Gear Fault Detection Research Articles

Anomaly Detection and Remaining Useful Life Estimation for the Health and Usage Monitoring Systems 2023 Data Challenge.

Gear fault detection and remaining useful life estimation are important tasks for monitoring the health of rotating machinery. In this study, a new benchmark for endurance gear vibration signals is presented and made publicly available. The new dataset was used in the HUMS 2023 conference data challenge to test anomaly detection algorithms. A survey of the suggested techniques is provided, demonstrating that traditional signal processing techniques interestingly outperform deep learning algorithms in this case. Of the 11 participating groups, only those that used traditional approaches achieved good results on most of the channels. Additionally, we introduce a signal processing anomaly detection algorithm and meticulously compare it to a standard deep learning anomaly detection algorithm using data from the HUMS 2023 challenge and simulated signals. The signal processing algorithm surpasses the deep learning algorithm on all tested channels and also on simulated data where there is an abundance of training data. Finally, we present a new digital twin that enables the estimation of the remaining useful life of the tested gear from the HUMS 2023 challenge.

Active learning for gear defect detection in gearboxes

Active learning for gear defect detection in gearboxes

A novel feature mode decomposition method and its application for gear fault detection

Fault detection can promptly reveal the potential hazards of mechanical equipment, guaranteeing the safety, stability, and reliability of their operation. Although many advanced fault detection methods, such as spectral kurtosis, deconvolution and decomposition, have been developed, most of them still suffer from insufficient utilization of fault features and incomplete extraction of diagnostic information. Given this, we propose a novel method named feature mode decomposition (FMD). Firstly, to coarsely steer the decomposition direction, a finite impulse response (FIR) filter bank is set up with a window initialization. The, correlated kurtosis (CK) is taken to evaluate the latent fault-related information in mode signals, thereby guides the updating process of all adaptive filters, assisted with period estimation. Ultimately, the unnecessary and intermingled modes are weeded out by mode selection. Experimental cases verified that the proposed FMD can adaptively decompose the fault mode of gear fault signal with excellent inspection ability. Comparison between the results and those of the classical variational mode decomposition (VMD) further highlights that FMD is more robust to other interference and noise.

Early pitting fault detection for polymer gears using kurtosis-VMD based condition indicators

The vibration signals of a polymer gear are considerably weak and susceptible to ambient noise at the early stage of the fault, which makes the fault difficult to detect. Efficient detection of an early fault in a polymer gear may improve the operation safety of the machinery system that utilizes it for power transmission. This study introduces an innovative approach for the early detection of pitting faults in polymer gears, utilizing condition indicators (CIs) derived from kurtosis-variational mode decomposition (VMD). First, the vibration signal of the polymer gear is decomposed using VMD into several components. Second, the sensitive components are selected to construct a new signal from the first two largest kurtosis values. Third, the CIs are extracted from newly constructed signals, and envelope spectrum analysis is performed. It is observed from the results that the kurtosis-VMD based CIs are effective in the early pitting fault detection of polymer gears. Finally, it is found that the proposed method performs better in all operating conditions considered in the experiment, compared with raw signal and kurtosis-empirical mode decomposition (EMD) based analysis. The proposed method’s response to noise is also explored. Furthermore, the proposed method is compared with the existing time synchronous averaging (TSA), difference, and residual methods.

An Effective Dataset Preprocessing Method in Tilted Gear Defects Target Detection

Gear defect detection is a crucial component in power automation systems. Methods based on deep learning have exhibited excellent performance in detecting gears. However, the effect of defect detection for tilted gears is not as good. This is due to the traditional horizontal bounding box annotation method, which inevitably produces many overlapping areas in the annotated bounding box when used in tilted gear targets, and a large part of the areas do not belong to defects. In order to address the issue of low precision in the detection of defects in tilted gears based on deep learning, a dataset preprocessing scheme has been proposed. Initially, the images and annotation files of the training set and validation set are automatically rotated. Subsequently, the rotated data are utilized to train the defect detection model. Finally, the test set is subjected to the same image rotation method as the first step, resulting in the generation of high‐precision defect detection results, which are then input into the defect detection model. In order to facilitate comparison with the ground truth, the JSON file obtained from the detection result is rotated in reverse and the result is mapped to the original test set before rotation. In order to verify the effectiveness of the dataset optimization method proposed in this article, the same configuration file is used to train and evaluate the gear defect detection model with the dataset before and after optimization, respectively. The three detection models with the highest comprehensive indicators were employed to assess the dataset before and after optimization, with the average detection precision mAP serving as a quantitative comparison indicator. The findings demonstrate that the data optimization method proposed in this article has markedly enhanced mAP in tilted gear defect detection. Upon testing the proposed method on the PP‐YOLOE + model, mAP (0.5) exhibited an increase of 18.5%, while mAP (0.5 : 0.95) demonstrated a 29.1% improvement. When the method was tested on the RT‐DETR model, mAP (0.5) exhibited an increase of 24.7%, while mAP (0.5 : 0.95) demonstrated an 18.4% improvement. When tested on the few‐short model, mAP (0.5) increased by 23.3% and mAP (0.5 : 0.95) increased by 19.8%. This validates the effectiveness of the dataset processing method proposed in this article in the detection of tilted gear defects.

Encoder Signal Analysis and its Application in Gear Fault Detection

Encoder Signal Analysis and its Application in Gear Fault Detection

Variational Mode Decomposition based Vibro-Acoustic Analysis for Spur Gear Fault Detection

Feature extraction from dynamic response characteristics of the machinery serves as an index for quantifying its health status. Vibration response analysis based on the spectral distribution of gear-mesh frequency is the industrial norm for gearbox fault detection. Investigations based on the efficacy of other responses, such as sound generated by the gearbox, are limited. The air pressure induced around the vicinity of the machinery may also indicate the health status. This paper studies the applicability of sound pressure signals as the basis for gear fault detection. The paper also investigates the efficacy of combined sound and vibration analysis for fault detection. The fault detection ability of the above-mentioned single-signal analysis and combined-signal analysis are compared with traditional vibration signal analysis. Variational Mode Decomposition and Multivariate Variational Mode Decomposition are used to extract noise-free signals centred around the required centre frequency in the case of single-signal analysis and a combined signal analysis, respectively. The study is conducted on an in-house test rig with a seeded gear tooth fault.

A New Automated Classification Framework for Gear Fault Diagnosis Using Fourier–Bessel Domain-Based Empirical Wavelet Transform

Gears are the most important parts of a rotary system, and they are used for mechanical power transmission. The health monitoring of such a system is needed to observe its effective and reliable working. An approach that is based on vibration is typically utilized while carrying out fault diagnostics on a gearbox. Using the Fourier–Bessel series expansion (FBSE) as the basis for an empirical wavelet transform (EWT), a novel automated technique has been proposed in this paper, with a combination of these two approaches, i.e., FBSE-EWT. To improve the frequency resolution, the current empirical wavelet transform will be reformed utilizing the FBSE technique. The proposed novel method includes the decomposition of different levels of gear crack vibration signals into narrow-band components (NBCs) or sub-bands. The Kruskal–Wallis test is utilized to choose the features that are statistically significant in order to separate them from the sub-bands. Three classifiers are used for fault classification, i.e., random forest, J48 decision tree classifiers, and multilayer perceptron function classifier. A comparative study has been performed between the existing EWT and the proposed novel methodology. It has been observed that the FBSE-EWT with a random forest classifier shows a better gear fault detection performance compared to the existing EWT.

Fault feature detection of planet gear in multi-stage planetary gearbox based on angle compensation synchronous averaging

Planetary gearboxes are often used for power transmission in wind turbines. It makes the gear fault detection of planetary gearboxes very useful and necessary. Generally, windowed synchronous averaging (WSA) is a widely used method to detect planet gear failure in planetary gearboxes. However, the WSA is not directly suitable for the multi-stage planetary gearbox since different planetary stages are working together and vibrations from different stages are coupled to each other. The coupled interference is the synchronous interference, which cannot be reduced by the WSA. It makes the faulty planet gear of interest difficult to detect. To address this issue, an angle compensation synchronous averaging (ACSA) scheme is proposed. The equi-angle resampling is applied to the obtained vibration first. Then, the synchronous interference can be constructed by applying the synchronous averaging-based angle compensation strategy, and removed by subtracting it from the resampled vibration. Subsequently, the fast spectral kurtosis is utilized on the residual vibration to obtain the envelope signal. Next, the WSA is further applied to the residual vibration and envelope signal for constructing the synthetic vibration and synthetic envelope signal corresponding to the planet gear of interest. Finally, the narrowband demodulation and envelope analysis are used to detect planet gear failure. Experiments on a two-stage planetary gearbox positively support the ACSA scheme.

Intelligent and Small Samples Gear Fault Detection Based on Wavelet Analysis and Improved CNN

Traditional methods for identifying gear faults typically require a substantial number of faulty samples, which in reality are challenging to obtain. To tackle this challenge, this paper introduces a sophisticated approach for intelligent gear fault identification, utilizing discrete wavelet decomposition and an enhanced convolutional neural network (CNN) optimized for scenarios with limited sample data. Initially, the features of the sample signal are extracted and enhanced using discrete wavelet decomposition. Subsequently, the refined signal is transformed into a two-dimensional image through a Markov transition field, preparing it for improved two-dimensional CNN training. Finally, the refined network model is applied to assess the gear fault dataset, achieving a training accuracy of 97% and a classification accuracy of 88.33%. This demonstrates the method’s feasibility and effectiveness in identifying gear faults with limited sample data.

Development of Non-Conventional Method of Lissajous pattern for gear fault diagnosis using Machine Learning Technique

Gears are major elements in machine parts and automotive systems, to transmitting torque and energy between components. As gears are under severe loads and stresses, they can create different types of faults and defects over period, impacting their own performance and reliability. Fault detection is an essential part of gear maintenance because that allows potential issues to be detected before they induce failures or expensive breakdowns. Technicians can identify early indications of wear, imbalance, damage, or even other concerns that might compromise their performance or lead to premature failure by monitoring the condition of gears. The industry has conventionally used vibration signals, thermography, oil analysis, and visually inspecting to diagnose faults. These techniques also have significant limitations which can decrease their accuracy in detecting faults and the cost is also very high. To overcome these limitations, new methods such as Lissajous pattern developed for fault detection in gears. These techniques have proved impressive outcomes in fault detection which have been previously undetectable using Conventional methods. This study evaluates the accuracy, performance, and cost-effectiveness of conventional and non-conventional techniques to determine the best method for fault diagnosis. Once compared to the conventional method, the Lissajous method obtained better accuracy than conventional method.

Spur Gear Fault Detection Using Design of Experiments and Support Vector Machine (SVM) Algorithm

Spur Gear Fault Detection Using Design of Experiments and Support Vector Machine (SVM) Algorithm

An industrial interference-resistant gear defect detection method through improved YOLOv5 network using attention mechanism and feature fusion

Automatically detecting the metallic gear defects may encounter the misdetected and undetected errors when the defect features are very close to the interferences of oil stains and image shadow. To address these challenges, in this paper an industrial interference-resistant gear surface defect detection method is presented based on the improved YOLOv5 network. The method combines the approach using the attention module CBAMC3 constituted by CBAM and C3 modules together with the analysis of the module BiFPN_concat, which makes the network capable of extracting and fusing defect features, respectively. The cosine annealing function is then employed to improve the learning ability of the network by modifying the learning rate. The experimental results indicate that the improved YOLOv5 network enhances both the recall rate by 13.1% and the mAP@0.5 by 12% compared with the original YOLOv5 network, and also outweighs the classical algorithms in detection speed with average 25 frames per second.

A spectral kurtosis based blind deconvolution approach for spur gear fault diagnosis

Unanticipated background noises often convolute fault information in the gearboxes’ vibration response. The Blind Deconvolution strategy has been extensively applied for fault-impulse enhancement to aid gear fault detection. The existing deconvolution strategies involve designing an optimum filter applied in the time domain. Gear tooth wear leads to the excitation of Gear Mesh Frequency harmonics. Hence, spectral analysis is typically used for gearbox fault detection. As such, feature enhancement in the order domain is more practical than existing blind deconvolution approaches. This study proposes a Spectral Kurtosis-based blind deconvolution strategy with filtering done in the order domain, to aid gear fault detection. Experimental analyses show 109.76% and 64.48% better performance for constant and real-world speed operation, respectively, for the proposed method to aid spectral analysis-based fault detection.

Comparing Reservoir Artificial and Spiking Neural Networks in Machine Fault Detection Tasks

For the last two decades, artificial neural networks (ANNs) of the third generation, also known as spiking neural networks (SNN), have remained a subject of interest for researchers. A significant difficulty for the practical application of SNNs is their poor suitability for von Neumann computer architecture, so many researchers are currently focusing on the development of alternative hardware. Nevertheless, today several experimental libraries implementing SNNs for conventional computers are available. In this paper, using the RCNet library, we compare the performance of reservoir computing architectures based on artificial and spiking neural networks. We explicitly show that, despite the higher execution time, SNNs can demonstrate outstanding classification accuracy in the case of complicated datasets, such as data from industrial sensors used for the fault detection of bearings and gears. For one of the test problems, namely, ball bearing diagnosis using an accelerometer, the accuracy of the classification using reservoir SNN almost reached 100%, while the reservoir ANN was able to achieve recognition accuracy up to only 61%. The results of the study clearly demonstrate the superiority and benefits of SNN classificators.

Rotor imbalance detection and diagnosis in floating wind turbines by means of drivetrain condition monitoring

This paper presents a novel approach for detection and diagnosis of the rotor imbalance types pitch misalignment, yaw misalignment and mass imbalance by monitoring the drivetrain vibration response. Traditionally, only SCADA signals including nacelle accelerations, rotor speed and electrical power are utilized for this purpose, while drivetrain condition monitoring signals are mainly used for fault detection in gears and bearings. A diagnostic method is proposed using statistical change detection methods for fault detection, phase angle estimation for localizing the faulty blade, and physics-based decision criteria for fault classification. The proposed method is tested in a numerical case study with aeroelastic and drivetrain multi-body models of the 10 MW DTU reference wind turbine. The results suggest that drivetrain condition monitoring signals are particularly beneficial for detecting and diagnosing pitch misalignment, since this fault type uniquely induces periodic out-of-plane bending moments that excite drivetrain bending modes. Drivetrain signals improved the detection rate of a 1° pitch error from 19% to near 100% and reduced the standard error in locating the faulty blade from 71.5° to 11.2°. In addition, by using drivetrain vibration amplitudes as a decision criterion, all considered pitch error cases are correctly distinguished from other fault types.

Defect detection of gear parts in virtual manufacturing

Gears play an important role in virtual manufacturing systems for digital twins; however, the image of gear tooth defects is difficult to acquire owing to its non-convex shape. In this study, a deep learning network is proposed to detect gear defects based on their point cloud representation. This approach mainly consists of three steps: (1) Various types of gear defects are classified into four cases (fracture, pitting, glue, and wear); A 3D gear dataset was constructed with 10000 instances following the aforementioned classification. (2) Gear-PCNet+ + introduces a novel Combinational Convolution Block, proposed based on the gear dataset for gear defect detection to effectively extract the local gear information and identify its complex topology; (3) Compared with other methods, experiments show that this method can achieve better recognition results for gear defects with higher efficiency and practicability.

Enhanced weight symplectic geometry decomposition based on maximum periodic kurtosis deconvolution

To enhance the periodic impact component and improve the accuracy of planetary gear fault detection, an enhanced weighted symplectic geometry decomposition based on maximum periodic kurtosis deconvolution (MPKD-EWSGD) is proposed in the paper. On the one hand, MPKD-EWSGD adopts the MPKD method for noise reduction preprocessing to highlight the periodic impulse component. On the other hand, MPKD-EWSGD introduces the periodic impulse intensity (PII) to choose components with fault information, avoiding the disadvantages of denoising methods that use the component energy as the measurement standard. Emulation and experimental signals show that MPKD-EWSGD can effectively reduce noise and is an effective method for planetary gearbox fault detection.

Supervised Machine Learning Based Approach for Early Fault Detection in Polymer Gears Using Vibration Signals

Supervised Machine Learning Based Approach for Early Fault Detection in Polymer Gears Using Vibration Signals

Gear fault detection in gearboxes operated in non-stationary conditions based on variable sideband analysis without a tachometer

Machine fault diagnostic techniques by vibration signal analysis have been applied and widely developed in the industry. In recent years, using the least number of measuring devices to simplify the diagnostic process of gear fault has attracted many researchers. Various digital signal analysis methods in the time and frequency domains have been effectively applied to detect gear faults of gearboxes operated in stationary conditions or rotational speeds with minor fluctuations. However, in reality, gearboxes often work at rotational speeds with large fluctuations and variable loads due to technological progress. In such cases, the measured vibration signal is the amplitude–frequency modulated signal. Therefore, the traditional vibration analysis methods have become ineffective. The modulated signal analysis techniques often require additional key phase information measured using a tachometer to achieve good diagnostic results. However, setting up the tachometer while operating the machine is not effortless. This paper offers a new approach combining many methods to detect gear faults based on sideband analysis in the time–frequency distribution of measured vibration signals using only a single measurement channel. This study is applied in many cases where the machine cannot stop and reduce instrumentation requirements and can develop to build an online diagnostic system.