Fault Diagnosis System Research Articles

Dynamic time warping based instantaneous frequency estimation for rotating machineries under non-stationary operating conditions

Abstract For tacho-less order tracking, it is crucial to achieve an accurate estimation of instantaneous phase and frequency under non-stationary conditions for fault diagnosis of mechanical systems. Therefore, an instantaneous frequency estimation method based on dynamic time warping for rotating machinery was proposed. Firstly, a mono-component signal is obtained by variational mode decomposition of multi-harmonic vibration signal generated in practical engineering. Then, a stationary reference signal is constructed by the mean frequency of the vibration signal, which is obtained by the signal spectrum. Aligning the vibration signal with the reference signal to obtain the rough phase estimation and warping path. Finally, a polynomial model is constructed to calculate the analytic expression of the smooth warping path to obtain the instantaneous frequency result. The simulation and experiment present that the dynamic time warping-based method can estimate the instantaneous frequency of the signal under non-stationary conditions. Compared with the classical Hilbert transform method, the proposed dynamic time-warping algorithm has stronger noise resistance.

Research on low-speed unmanned vehicle diagnosis system based on fault tree

Abstract To reduce the safety risks of low-speed unmanned vehicles losing control and colliding with pedestrians or vehicles due to malfunctions during autonomous task execution, Fault Tree Analysis (FTA) is used to analyze the safety of low-speed unmanned vehicles. A low-speed unmanned vehicle fault diagnosis system is designed to take corresponding measures in case of vehicle faults and improve vehicle driving safety.

Intelligent fault diagnosis for unbalanced battery data using adversarial domain expansion and enhanced stochastic configuration networks

An accurate and efficient fault diagnosis method for battery systems is crucial to ensuring the safety of battery packs. Addressing the issue of insufficient actual fault data in battery operations, this paper proposes an intelligent fault diagnosis method based on feature-enhanced stochastic configuration networks and adversarial domain expansion of imbalanced battery fault data (AFDEM-FESCN). Firstly, we designed an adversarial fault domain data expansion method (AFDEM). By learning the distribution of fault data through adversarial training, the distribution of sample domains is balanced, thereby reducing model bias. Subsequently, we adjusted the distribution of SCN iterative parameters and added a linear feature layer. This enhances the feature extraction capability of the network through distribution overlay, enabling fault diagnosis. Finally, the effectiveness and feasibility of the proposed method were validated through a practical battery system fault diagnosis case, achieving a diagnostic accuracy of 92.1%. Experimental results demonstrate that the AFDEM-FESCN method exhibits good accuracy in battery system fault diagnosis, providing an effective solution to the challenge of imbalanced data in intelligent fault diagnosis.

Improvement of local outlier factor algorithms for lithium-ion battery fault diagnosis

Fault diagnosis is one of the most important active strategies to protect lithium-ion batteries (LIBs) from safety accidents. The tasks of fault diagnosis usually can be divided into three levels, i.e., (1) fault detection, (2) fault isolation, and (3) fault estimation. The local outlier factor (LOF) method has been proved effective in conducting fault detection (level 1 of fault diagnosis) for LIB energy storage systems (ESSs). However, the original LOF algorithm (OLOF) may fail to make correct fault estimation (level 3 of fault diagnosis). Therefore, in this study, two novel LOF algorithms, i.e., the improved LOF algorithm (ILOF) and the simplified LOF algorithm (SLOF), are put forward to enhance the capability of fault estimation. The performance of the three LOF algorithms: OLOF, ILOF and SLOF, are firstly compared via three mathematical cases. Their capabilities in fault diagnosis are further investigated based on the simulated data from an air-cooled LIB ESS and the experimental data from a water-cooled LIB ESS. Results prove the effectiveness of all the three algorithms in fault detection; nevertheless, when the OLOF algorithm fails to make correct fault estimation, both of the two proposed novel algorithms, i.e., the ILOF algorithm and the SLOF algorithm, present the capability in making correct fault estimation. It is found as well that the ILOF algorithm is better than the SLOF algorithm in robustness though the latter is more efficient than the former in memory usage and computational time.

A Cost-Effective Fault Diagnosis and Localization Approach for Utility-Scale PV Systems Using Limited Number of Sensors

A Cost-Effective Fault Diagnosis and Localization Approach for Utility-Scale PV Systems Using Limited Number of Sensors

Modelling of electromechanical coupling dynamics for high-speed EHT system used in HEV and characteristics analysis

As the development of hybrid electric vehicle (HEV) to high-speed, heavy-load, the electromechanical coupling vibration becomes the bottleneck in high-speed electromechanical hybrid transmission (EHT) system. To explore the dynamics characteristics and optimization direction for high-speed EHT system, an electromechanical coupling dynamics model under multi-source excitations is established with lumped-distributed parameter method and verified with experiment data. The dynamics model shows higher accuracy in both time and frequency domain analysis. On this basis, firstly, the comparison between lumped-shaft and distributed-shaft model is studied under time and frequency domain. Comparing with the lumped-shaft model, the distributed-shaft model shows higher accuracy, and can better reflect the coupling vibration of multi-stage planetary gears (PGs). Secondly, the inherent vibration model is derived, and the effect of electromechanical coupling and high-speed working conditions on inherent vibration characteristics are studied. Thirdly, a new method called ‘machine-electricity-magnet coupling interface’ is proposed to reveal the coupling vibration phenomenon. In addition, the signal of stator current and electromagnetic torque includes the frequency of PGs, which is a basis on the fault diagnosis and state monitor of EHT system. Last but not the least, the vibration acceleration of EHT system is analysed under variable speed and load working conditions. Rotation speed and gear meshing force is found as the main influencing factor of series EHT system, and the specific optimization direction is also given.

Toward understandable semi-supervised learning fault diagnosis of chemical processes based on long short-term memory ladder autoencoder (LSTM-LAE) and self-attention (SA)

Fault diagnosis and localization play vital role in chemical process monitoring. Finding the root cause of fault accurately and timely is the key to avoid serious accidents and ensure process safety. Unfortunately, most deep learning-based models only involves in predicting fault state and interpretability is still not fully explored. Besides, lack of labeled samples in practical situations makes supervised learning difficult to implement. To circumvent the obstacle, semi-supervised learning based on long short-term memory ladder autoencoder is combined with self-attention mechanism, which is intended to establish interpretable model by explicitly clarifying the corresponding relationship between abnormal variables and faults. Using Tennessee Eastman process and practical high-purity carbonate production process as benchmark, fault diagnosis and identification performance of proposed LSTM-LAE-SA is validated, and interpretability analysis is performed to demonstrate its capabilities in abnormal variable locations. The developed understandable model could improve operator's trust and industrial application in fault diagnosis system.

A novel intelligent multicross domain fault diagnosis of servo motor-bearing system based on Domain Generalized Graph Convolution Autoencoder

A novel intelligent multicross domain fault diagnosis of servo motor-bearing system based on Domain Generalized Graph Convolution Autoencoder

Empirical study on fine-tuning pre-trained large language models for fault diagnosis of complex systems

With increasing complexity and interconnectivity of the modern industrial systems, effectively diagnosing faults is a core step to Prognostic and Health Management (PHM). Language models, particularly Large Language Models (LLMs) that pre-trained on massive corpora, have demonstrated remarkable capabilities in Natural Language Processing (NLP) and its related downstream tasks. However, how to leverage these models for facilitating system fault diagnosis and verify its effectiveness is rarely explored. This paper incorporates the potential comprehension ability of pre-trained LLMs, and investigates the efficacy of fine-tuning LLMs for realizing efficient system fault diagnosis. The experiments conducted in this study involve both open-source and closed-source models, and utilize a simulation and a real fault diagnosis dataset. We find that these models consistently achieve high performance across various metrics compared to the baselines. Additionally, qualitative and quantitative analysis is performed to investigate several aspects of the approach, such as the impact of dataset size, data normalization, missing values and explainability of the diagnosis, further showcasing the potential as well as limitations of the approach.

Power system fault diagnosis with quantum computing and efficient gate decomposition

Power system fault diagnosis is crucial for identifying the location and causes of faults and providing decision-making support for power dispatchers. However, most classical methods suffer from significant time-consuming, memory overhead, and computational complexity issues as the scale of the power system concerned increases. With rapid development of quantum computing technology, the combinatorial optimization method based on quantum computing has shown certain advantages in computational time over existing methods. Given this background, this paper proposes a quantum computing based power system fault diagnosis method with the quantum approximate optimization algorithm. The proposed method reformulates the fault diagnosis problem as a Hamiltonian by using Ising model, which completely preserves the coupling relationship between faulty components and various operations of protective relays and circuit breakers. Additionally, to enhance problem-solving efficiency under current equipment limitations, the symmetric equivalent decomposition method of multi-z-rotation gate is utilized. Furthermore, the small probability characteristics of power system events is utilized to reduce the number of qubits. Simulation results based on the test system show that the proposed methods can achieve the same optimal results with a faster speed compared with the classical higher-order solver provided by D-Wave.

A fault diagnosis method of hot spots for photovoltaic clusters based on model parameters

AbstractUtilizing the direct current‐side electrical data resources of the photovoltaic power generation system on the FusionSolar platform, this study investigates the impact of hot spot faults on the output characteristics of photovoltaic strings and proposes a hot spot fault diagnosis method based on time series waveform characteristics. By analyzing the mechanisms of hot spot generation and evolution, as well as the characteristic differences in I–V curves and time series compared to other types of faults, the waveform variation patterns of hot spots in current and voltage time series are obtained. A function form suitable for hot spot fault waveform characteristics in time series graphs is constructed, and fault diagnosis feature vectors are extracted. Combining field operation and maintenance experience, a fuzzy reasoning fault diagnosis system is established to determine the causes and estimate the severity of hot spot faults. Experimental results indicate that hot spot faults have unique and corresponding variations in the current/voltage time series waveforms of the string output. The constructed function form can clearly represent the waveform variation patterns, and the established fuzzy reasoning system can achieve effective and reliable diagnosis of hot spot faults.

Fault Diagnosis Method for Tractor Transmission System Based on Improved Convolutional Neural Network–Bidirectional Long Short-Term Memory

In response to the problems of limited algorithms and low diagnostic accuracy for fault diagnosis in large tractor transmission systems, as well as the high noise levels in tractor working environments, a defect detection approach for tractor transmission systems is proposed using an enhanced convolutional neural network (CNN) and a bidirectional long short-term memory neural network (BILSTM). This approach uses a one-dimensional convolutional neural network (1DCNN) to create three feature extractors of varying scales, directly extracting feature information from different levels of the raw vibration signals. Simultaneously, in order to enhance the model’s predicted accuracy and learn the data features more effectively, it presents the multi-head attention mechanism (MHA). To overcome the issue of high noise levels in tractor working environments and enhance the model’s robustness, an adaptive soft threshold is introduced. Finally, to recognize and classify faults, the fused feature data are fed into a classifier made up of bidirectional long short-term memory (BILSTM) and fully linked layers. The analytical findings demonstrate that the fault recognition accuracy of the method described in this article is over 98%, and it also has better performance in noisy environments.

SCAE-Stacked Convolutional Autoencoder for Fault Diagnosis of a Hydraulic Piston Pump with Limited Data Samples.

Deep learning (DL) models require enormous amounts of data to produce reliable diagnosis results. The superiority of DL models over traditional machine learning (ML) methods in terms of feature extraction, feature dimension reduction, and diagnosis performance has been shown in various studies of fault diagnosis systems. However, data acquisition can sometimes be compromised by sensor issues, resulting in limited data samples. In this study, we propose a novel DL model based on a stacked convolutional autoencoder (SCAE) to address the challenge of limited data. The innovation of the SCAE model lies in its ability to enhance gradient information flow and extract richer hierarchical features, leading to superior diagnostic performance even with limited and noisy data samples. This article describes the development of a fault diagnosis method for a hydraulic piston pump using time-frequency visual pattern recognition. The proposed SCAE model has been evaluated on limited data samples of a hydraulic piston pump. The findings of the experiment demonstrate that the suggested approach can achieve excellent diagnostic performance with over 99.5% accuracy. Additionally, the SCAE model has outperformed traditional DL models such as deep neural networks (DNN), standard stacked sparse autoencoders (SSAE), and convolutional neural networks (CNN) in terms of diagnosis performance. Furthermore, the proposed model demonstrates robust performance under noisy data conditions, further highlighting its effectiveness and reliability.

Fault diagnosis of drivetrain systems based on vibration signals and convolutional neural networks

Accurate and fast fault diagnosis can improve the reliability of drivetrain systems. The fault diagnosis model based on convolutional neural networks is tricky because the accuracy and efficiency of diagnosis depend on the neural network parameters, such as layer size and number of layers. In this paper, we developed a data pre-processing technique to format vibrational data to be receptible to a neural network. First one will study the difference in performance by inputting a spectrogram of the vibrational data compared to the raw data itself. Second, the windowizing technique is used and its parameters are explored to train different configurations of convolutional neural networks. Finally, the optimal neural network parameters are obtained to improve the computational efficiency while the training model has the highest accuracy.

Event-Triggered Collaborative Fault Diagnosis for UAV–UGV Systems

The heterogeneous unmanned system, which is composed of unmanned aerial vehicles (UAV) and unmanned ground vehicles (UGV), has been broadly applied in many domains. Collaborative fault diagnosis (CFD) among UAVs and UGVs has become a key technology in these unmanned systems. However, collaborative fault diagnosis in unmanned systems faces the challenges of the dynamic environment and limited communication bandwidth. This paper proposes an event-triggered collaborative fault diagnosis framework for the UAV–UGV system. The framework aims to achieve autonomous fault monitoring and cooperative diagnosis among unmanned systems, thus enhancing system security and reliability. Firstly, we propose a fault trigger mechanism based on broad learning systems (BLS), which utilizes sensor data to accurately detect and identify faults. Then, under the dynamic event triggering mechanism, the network communication topology between the UAV–UGV system and BLS is used to achieve cooperative fault diagnosis. To validate the effectiveness of our proposed scheme, we conduct experiments on a software-in-the-loop (SIL) simulation platform. The experimental results demonstrate that our method achieves high diagnosis accuracy for the UAV–UGV system.

Development of an acoustic fault diagnosis system for UAV propeller blades

With the rapid growth in demand for unmanned aerial vehicles (UAVs), novel maintenance technologies are essential for ensuring automatic, safe, and reliable operations. This study compares two fault detection systems that utilize the acoustic signature of UAV propeller blades for classifying their health state. By employing an acoustic camera with 112 microphones for spatial resolution of sound sources, datasets of acoustic images are generated in three differently reverberating environments for the third octave frequency bands of 6300 Hz, 8000 Hz, 10,000 Hz and 12,500 Hz. A convolutional neural network (CNN) is trained and evaluated with maximum F1-scores of 0.9962 and 0.9745 for two and three propeller health classes, respectively. Furthermore, we propose a second approach based on a linear classification (LC), which utilizes a rotating beamformer for comparison. This approach uses only two sound sources that are identified after the acoustic beamforming of a two-bladed propeller. In comparison, this algorithm detects propeller tip damages without applying a machine learning algorithm and reaches a slightly lower F1-score of 0.9441.

SS-CWGAN: A novel fault diagnosis model for building HVAC systems under limited labeled data

Fault diagnosis (FD) systems are important to the reliable operations and energy management of the building HVAC systems. However, traditional supervised learning-based FD methods heavily rely on labeled data and underutilize massive unlabeled data. This study proposes a semi-supervised conditional Wasserstein generative adversarial network (SS-CWGAN) for the FD of HVAC systems under limited labeled data. The information contained in a large number of unlabeled data is fully leveraged to improve the fault diagnosis performance of HVAC systems. Besides, a unique semi-supervised learning approach is developed for SS-CWGAN to extract the common distribution information of labeled and unlabeled data, enhancing the model generalization, especially in the practical scenario with limited labeled data. Comparison experiments demonstrate that when the number of labeled data is within 100, the proposed method outperforms the existing supervised and semi-supervised FD methods by more than 14% and 4% in diagnosis accuracy, respectively. Furthermore, the diagnosis performance improvement of the proposed method is more significant at severity level 1, suggesting the promising ability to diagnose the faults at an early stage.

Research on Key Technology of Wind Turbine Drive Train Fault Diagnosis System Based on Digital Twin

Fault diagnosis of wind turbines has always been a challenging problem due to their complexity and harsh working conditions. Although data-mining-based fault diagnosis methods can accurately and efficiently diagnose potential faults, the visibility is extremely poor. In this paper, digital twin technology is introduced into the fault diagnosis of wind turbine drive train systems, and a wind turbine drive train fault diagnosis method based on digital twin technology is proposed, which monitors and simulates the actual operating condition in real-time by establishing a digital twin model of the wind turbine drive train. In addition, an improved variational modal decomposition combined with particle swarm optimization least squares support vector machine (IVMD-PSO-LSSVM) fault diagnosis method is proposed, which not only improves the accuracy of fault diagnosis but also effectively shortens the diagnosis time and strengthens the response speed of the system. Finally, a digital twin system for condition monitoring and fault diagnosis of wind turbine drive trains is developed based on the Unity 3D platform. Experiments show that the proposed IVMD-PSO-LSSVM can accurately identify fault types with an accuracy rate of 99.1%, which is an improvement of 3.4% compared to before. The proposed digital twin model can be used for real-time monitoring of wind turbine vibration data and provide a more intuitive real-time simulation of the wind turbine’s operating status. This facilitates quick fault location and enables more accurate and efficient maintenance.

Hydraulic system fault diagnosis decoupling method based on 2D time-series modeling and self-attention fusion

Hydraulic systems play a pivotal and extensive role in mechanics and energy. However, the performance of intelligent fault diagnosis models for multiple components is often hindered by the complexity, variability, strong hermeticity, intricate structures, and fault concealment in real-world conditions. This study proposes a new approach for hydraulic fault diagnosis that leverages 2D temporal modeling and attention mechanisms for decoupling compound faults and extracting features from multisample rate sensor data. Initially, to address the issue of oversampling in some high-frequency sensors within the dataset, variable frequency data sampling is employed during the data preprocessing stage to resample redundant data. Subsequently, two-dimensional convolution simultaneously captures both the instantaneous and long-term features of the sensor signals for the coupling signals of hydraulic system sensors. Lastly, to address the challenge of feature fusion with multisample rate sensor data, where direct merging of features through maximum or average pooling might dilute crucial information, a feature fusion and decoupling method based on a probabilistic sparse self-attention mechanism is designed, avoiding the issue of long-tail distribution in multisample rate sensor data. Experimental validation showed that the proposed model can effectively utilize samples to achieve accurate fault decoupling and classification for different components, achieving a diagnostic accuracy exceeding 97% and demonstrating robust performance in hydraulic system fault diagnosis under noise conditions.

Reinforced convolutional neural network fault diagnosis of industrial production systems

A new fault diagnosis method is proposed combining multiple models into a multiscale residual jagged dilated convolutional neural network with long short-term memory (MRJDCNN-LSTM). This method has strong feature extraction ability and can effectively extract and classify fault features. First, a new multi-scale residual jagged dilated convolution neural network (MRJDCNN) model is designed by incorporating the principle of residual learning into the improved jagged dilated convolution, and then it is combined with the long short term memory(LSTM) network delicately. This not only greatly improves the extraction rate of nonlinear high-latitude spatial features at different scales while preventing model degradation, but also subtly extracts time-related features, effectively reducing the loss of necessary features, and thus greatly improving the accuracy of model diagnosis. The simulation and comparative experiments of Tennessee-Eastman (TE) process and industrial coking furnace both prove that the constructed model has excellent performance.