Fault Diagnostics Research Articles

Fault diagnostics on railway switches and crossings using a bat-optimized sequence-to-sequence neural model

Fault diagnostics on railway switches and crossings using a bat-optimized sequence-to-sequence neural model

Fault diagnostics on railway switches and crossings using a bat-optimised sequence-to-sequence neural model

Fault diagnostics on railway switches and crossings using a bat-optimised sequence-to-sequence neural model

ИСПОЛЬЗОВАНИЕ АЛГОРИТМОВ МАШИННОГО ОБУЧЕНИЯ ДЛЯ ОБНАРУЖЕНИЯ И ДИАГНОСТИКИ НЕИСПРАВНОСТЕЙ В СИСТЕМАХ ЖИЗНЕОБЕСПЕЧЕНИЯ

The increasing interest in integrating machine learning methods into life support systems management is driven by their potential to enhance error detection, diagnostics, and forecasting capabilities. This paper focuses on analyzing current research to identify the strengths and weaknesses of machine learning models in the field of fault detection and prediction in building life support systems. It aims to provide insights into addressing issues such as data quality, algorithmic complexity, and interpretability. A review of contemporary scientific articles in this area was conducted, analyzing innovations and the practical implementation of machine learning models. The primary goal was to determine the optimal models for various scenarios, considering data characteristics, system complexity, task requirements, and desired outcomes. The information analysis led to formulating recommendations for integrating machine learning methods into life support systems, including the use of clustering methods for unlabeled data, XGBoost for handling unbalanced labeled data, support vector machines (SVM) for binary classification tasks, and decision trees (DT) or random forests (RF) for multi-class classification. Additionally, artificial neural networks (ANN) can be used for processing nonlinear data distributions. These recommendations are practically significant for enhancing the performance and energy efficiency of life support systems through the effective use of machine learning technologies. By providing guidance on model selection and implementation strategies, this work contributes to researchers, practitioners, and industry representatives in fault detection and diagnostics within life support systems.

Диагностика электротехнических компонентов гидротехнического оборудования: эффективность комплексного подхода

Diagnostics of electrical components of hydrotechnical equipment used in agricultural production within the planned maintenance intervals with continuous analysis of residual resource is useful for reducing equipment operation costs. The use of inexpensive universal diagnostic equipment in conjunction with primary analysis of the state of units makes it possible to detect 60 to 80% of pre-failure states. To ensure efficient application of complex approach to primary diagnostics of electrical components of hydraulic equipment used in agricultural production, the authors considered the most common electric motors of pumping units, electrohydraulic distributors and switching equipment. Complex diagnostics included the use of a thermal imager, vibration signal processing systems and current signal control systems. The vibrograms of equipment control points were analyzed, the state of bearing units, cavitation characteristics of pumps were assessed by methods of acoustic and vibration diagnostics, hydraulic equipment leaks were searched, power cables breakdown zones were detected in accordance with methodical instructions for determining the breakdowns of power cables with voltage up to 10 kV RD34.20.516‑90. The article shows the possibilities of applying the diagnostics methodology based on surface thermograms to determine typical defects. Particularly discussed are the parameters of current signals from the electromagnet coils observed in case of distribution equipment malfunction. The authors give comparative evaluation of diagnostics methods with calculated efficiency determination. It is established that the method of combined diagnostics using the analysis of external thermal fields can halve the time of preliminary diagnostics of faults with a stable forecast of the failure time 2‑3 months before the critical state. This reduces the costs for detection of unit failures and simplifies the planning of maintenance procedures.

A Gated Recurrent Generative Transfer Learning Network for Fault Diagnostics Considering Imbalanced Data and Variable Working Conditions.

Transfer learning (TL) and generative adversarial networks (GANs) have been widely applied to intelligent fault diagnosis under imbalanced data and different working conditions. However, the existing data synthesis methods focus on the overall distribution alignment between the generated data and real data, and ignore the fault-sensitive features in the time domain, which results in losing convincing temporal information for the generated signal. For this reason, a novel gated recurrent generative TL network (GRGTLN) is proposed. First, a smooth conditional matrix-based gated recurrent generator is proposed to extend the imbalanced dataset. It can adaptively increase the attention of fault-sensitive features in the generated sequence. Wasserstein distance (WD) is introduced to enhance the construction of mapping relationships to promote data generation ability and transfer performance of the fault diagnosis model. Then, an iterative "generation-transfer" co-training strategy is developed for continuous parallel training of the model and the parameter optimization. Finally, comprehensive case studies demonstrate that GRGTLN can generate high-quality data and achieve satisfactory cross-domain diagnosis accuracy.

Preface

2023 4th International Conference on Mechanical Engineering and Materials (ICMEM 2023) has been successfully held on November 2-4, 2023, in Wuhan China. Organized by Hubei Zhongke Institute of Geology and Environment Technology, ICMEM 2023 intends to invite worldwide famous scientists, experts, scholars, and researchers for academic presentations.ICMEM 2023 proceeding is a collection of outstanding submissions in the field of Dynamics, Vibration and Sound Engineering, Materials Science and Technology, Fluids Engineering, Solid Mechanics and Design, Micro/Nano Engineering and Technology, Production Engineering, Robotics and Control, Thermal and Power Engineering, Vehicle Technology, Machine Fault Diagnostics and Prognostics, Intelligent Transport System, Intelligent Monitoring Technology, Innovative Industrial Design, etc. Around 100 articles were selected from more than 220 submissions after peer-review, and authors and guests participated in ICMEM 2023 with keynote speeches, oral presentations and posters, promoting the communication among researchers and experts from institutes and universities, including National University of Singapore, Polytechnic University of Milan, Wuhan University, China University of Geosciences (Wuhan), Shenyang Aerospace University, Okayama University, Huazhong University of Science and Technology, Dongguan University Of Technology, etc.Cordial gratitude is delivered to Technical Program Committee of ICMEM 2023 for the professional work on guarantying the high standard of conference as well as the proceedings. We believe the proceedings will enhance the exchange of ideas and state-of-art knowledge on Mechanical Engineering and Materials.Professor GUPTA Manoj National University of Singapore, SingaporeAssociate Professor Jinyang Xu Shanghai Jiao Tong University, ChinaProfessor Krishna Murari Pandey National Institute of Technology Silchar, IndiaList of Technical Committee is available in this pdf.

Telematics and machine learning system for estimating the load condition of a heavy-duty vehicle

This study employs machine learning techniques to determine whether trucks are loaded or empty based solely on data from the vehicle's communication network, especially from the engine module. We achieved an accuracy of over 85% for short hauls (0.5 to 5 km) and nearly 95% for long hauls (5 to 500 km). This approach not only streamlines fleet management, reducing the need for communication between managers and drivers but also plays a crucial role in research related to fuel consumption reduction and intelligent fault diagnostics. Minimizing dependence on imprecise and time-delayed human responses opens new avenues for effective studies and innovative solutions in the transportation industry.

Implementing a Digital Twin-based fault detection and diagnosis approach for optimal operation and maintenance of urban distributed solar photovoltaics

Over the next decades, solar energy power generation is anticipated to gain popularity because of the current energy and climate problems and ultimately become a crucial part of urban infrastructure. Furthermore, in order to facilitate the rapid deployment of solar production in sustainable cities, solar technologies must be digitalized and smart. This will increase their reliability and ensure their optimal performance. This work fits into this perspective and focuses on the effectiveness of processing data using the Digital Twin (DT) framework as a state-of-art information technology for urban distributed solar PV systems. Accordingly, the present paper deals with the DT of five Building-added Photovoltaic (BAPV) systems installed at the solar village in Benguerir city, Morocco. The 3D models were generated in the Rhinoceros platform, the solar AC electrical production of each photovoltaic system is calculated by the ladybug-plugin-integrated statistical model, within the grasshopper environment. Additionally, data assessment and visualization are processed through Python. Then, a hybrid model-based and data-driven fault detection and diagnosis (FDD) approach is proposed to identify and isolate anomalies for decentralized solar PV systems at the urban scale using monitoring and inspection techniques, namely Remote Sensors (RS) and real-time solar production monitoring system. The application of the DT concept for complex dynamic systems has shown its effectiveness in ensuring optimal operating conditions for the energy systems by measuring the spatiotemporal energy performance of distributed solar PV systems at an urban scale, evolving with the changing environment, and identifying degradation based on anomalies occurrence; Only then when the physical models need to be amended. Therefore, the main goal behind this work is to advance knowledge regarding research on condition monitoring of solar systems in cities while considering issues related to anomalies like the degradation phenomenon in PV modules over time. The findings of this study have proved that the state of health and monitoring of PV systems could be successfully detected and efficiently monitored using the combination of real time monitoring and remote aerial sensing.

Innovative fault diagnosis and early warning method based on multifeature fusion model for electric vehicles

Fault diagnostics and early warning are crucial to the safe operation of lithium-ion batteries, and despite partial progress, it is still extremely difficult to solve the problem in a high-dimensional parameter space with complex failure causes. In this article, an innovative fault diagnosis and early warning method based on multifeature fusion model is designed for quantitative and qualitative comprehensive analysis and evaluation of the battery operating state information and the complex internal safety evolution trajectory. Firstly, to form a multi-dimensional fault feature matrix, a real vehicle dataset consisting of 400 vehicles is constructed using typical vehicle fault data from the cloud platform. Then, the original data is noise-reduced and pre-processed by applying initial bias correction and anomalous signal identification methods. Moreover, based on the equivalent circuit model and data-driven algorithm, 13 features including internal short circuit internal resistance, inconsistency accumulation risk, and principal element extraction index are designed from three dimensions: threshold, statistics and model. Finally, the multiple features are combined with random forest (RF) algorithm for hybrid fusion decision to achieve better generalization ability and suppress the overfitting risk of the algorithm. The results show that the proposed feature extraction and fusion decision methods can identify abnormal states and hazard levels in a timely and accurate manner, and this RF-based classification, warning and evaluation framework shows the promise of machine learning algorithms for interpretable early warning studies of battery failures, which can use more modal features to assess the hazard levels of actual operating battery systems in end-cloud collaboration scenarios.

Fault detection and diagnosis for variable refrigerant flow systems by using virtual sensors and deep learning

The primary objective of this work was to implement and evaluate a fault detection and diagnostic system for variable refrigerant flow (VRF) systems with multiple indoor units. To achieve this goal, virtual sensors and deep learning models are developed to predict the system performance whose direct measurement could be expensive or difficult. The primary challenges in implementing a diagnostic and performance monitoring system are the high initial costs of additional physical sensors. However, the combination of virtual sensor and the deep learning model can identify and isolate a condenser fouling fault. The virtual sensors were used to determine the condenser air flow rate in and the performance of a VRF system equipped with a variable-speed compressor and fans. An important step in automated fault detection and diagnostics is evaluating whether a condenser fouling fault is adequately significant to warrant service based on various criteria, including the potential to degrade comfort, increase in energy usage that can affect the overall system performance operation, and operating conditions that can shorten VRF system life. The deep learning model is developed to determine the performance degradation ratio between the performance estimated using the outputs of virtual sensors and that estimated using deep learning-based expected performance models. The fault impact models can help determine whether a fault is sufficiently serious to warrant a service recommendation.

Labelled dataset for Ultra-Low Temperature Freezer to aid dynamic modelling & fault detection and diagnostics

Ultra-low temperature (ULT) freezers are used to store perishable biological contents and are among the most energy-intensive equipment in laboratory buildings, biobanks, and similar settings. To ensure reliable and efficient operation, it is essential to implement data-driven fault detection and diagnostic algorithms, along with energy optimization techniques. This study presents labelled and long-term ULT-freezer performance dataset, the first of its kind, derived from 53 ULT freezers featuring two different control strategies. The dataset comprises high-resolution historical operation data spanning up to 10 years. More than 10 attributes are recorded from the freezing chamber and critical locations in the refrigeration systems. The dataset is labelled with regular events, such as door openings, as well as fault events obtained from 46 service reports. A scalable data pipeline, consisting of extraction, transformation, and loading processes, is developed to convert the raw data into a format ready for analysis. The dataset can be utilized to support the development of data-driven models and algorithms that advance the intelligent digital operation of ULT freezers.

Real-time intelligent fault diagnosis of rotating machines based on Archimedes algorithm optimised Gradient Boosting

ABSTRACT Intelligent fault diagnostics is an essential part of the concept of Industry 4.0. Most of the developed techniques are usually tested by the same dataset whose data is acquired under stable conditions and not generally assessed in terms of essential metrics that set the robustness and real-time suitability of a model. Based on such drawbacks, this study aims to propose an effective fault diagnosis technique to be employable in real-world settings. For this purpose, numerous ensemble learning methods and signal processing techniques are utilized and compared to find the ultimate fault analysis model suitable for real-time settings. The best approach is optimized by employing Archimedes Optimisation Algorithm (ArchOA) to eliminate the redundant features and therefore, reduce the computational load and time consumption. It is concluded that the Gradient Boosting Decision Tree (GBDT) is the most accurate model for all fault types and the most suitable one for potential real-time analysis considering its low time consumption and model-built time. Employing ArchOA reduces time consumption by 25% by eliminating redundant features by 36%. The proposed approach successfully identified the faulty state including the fault type with an average accuracy of 99.16% for the testing set and 97.50% for the real-time analysis.

A lightweight GAN-based fault diagnosis method based on knowledge distillation and deep transfer learning

Generative adversarial networks (GANs) have shown promise in the field of small sample fault diagnosis. However, it is worth noting that generating synthetic data using GANs is time-consuming, and synthetic data cannot fully replace real data. To expedite the GAN-based fault diagnostics process, this paper proposes a hybrid lightweight method for compressing GAN parameters. First, three modules are constructed: a teacher generator, a teacher discriminator, and a student generator, based on the knowledge distillation GAN (KD-GAN) approach. The distillation operation is applied to both teacher generator and student generator, while adversarial training is conducted for the teacher generator and the teacher discriminator. Furthermore, a joint loss function is proposed to update the parameters of the student generator by combining distillation loss and adversarial loss. Additionally, the proposed KD-GAN method is combined with deep transfer learning (DTL) and leverages real data to enhance the diagnostic model’s performance. Two numerical experiments are performed to demonstrate that the proposed KD-GAN-DTL method outperforms other GAN-based fault diagnosis methods in terms of computational time and diagnostic accuracy.

Deployment of real-time building automation system-integrated inverse-model-based fault detection and diagnostics algorithms

The complex operation of HVAC systems in large commercial buildings warrants regular implementation of advanced analytical approaches to operations and maintenance, and subsequent corrective measures to improve and maintain optimal energy performance. Despite the established capabilities of data-driven fault detection and diagnostics (FDD) to identify suboptimal controls policies and mechanical faults resulting in poor energy performance, few attempts have been made to deploy scalable solutions around these approaches. Furthermore, real-time BAS-integrated FDD methods are predominantly rule-based, offering limited insights to faults with gradual negative impacts to energy performance. This paper demonstrates the application of various established data-driven, inverse-model-based FDD methodologies in a BAS-integrated environment. Traditionally implemented sparingly, the novelty of recursive and automatic execution of advanced FDD methodologies, facilitated through a direct data pipeline to an existing BAS, capitalizes on the BAS’s real-time monitoring capabilities to enable continuously refreshed inverse model generation that can capture the gradual degradation of building performance, and provide up-to-date actionable visualizations and key performance indicators (KPI) to building operators. Since deployment, the application has successfully identified a scheduling fault on two separate occasions in a case study building in Ottawa, Canada, and the visualizations were presented to the building operators who resolved the issues.

A new method to perform Lithium-ion battery pack fault diagnostics – Part 2: Algorithm performance in real-world scenarios and cell-to-cell transferability

The last decade has witnessed significant progress towards zero-emission electric aviation. Lithium-ion batteries are at the centre of this technological transformation. However, electric flight requires suitable considerations for safety concerns associated with Lithium-ion batteries. Fault diagnostic approaches aimed at validating the safety of batteries before every flight have emerged as an attractive prospect. In this pursuit, a three-paper series proposing a novel fault diagnosis algorithm is presented. The algorithm is capable of diagnosing faults in an aircraft battery using the data collected during charging and was previously validated for a particular cell type under steady charging conditions. In this paper, two extensional aspects of the algorithm are investigated: cell-to-cell transferability and disparate charging conditions. While crucial for practical implementation, these aspects were overlooked in previous literature. Through experiments conducted at module-level, it was revealed that the algorithm could diagnose faults for a different cell type with minimal preliminary cell-level characterisation, thus demonstrating its ease of transferability. Moreover, modifications introduced in the algorithm enabled it to perform fault diagnosis under unsteady charging conditions for an aerospace module despite significant internal temperature gradients. Thus, the successful working of the algorithm in the considered aspects, which are unprecedented in literature, prove its feasibility for real-life application.

Augmented-State-Observer-Based Diagnostics of Open-Circuit and Sensor Faults in DFIG Wind Turbines

Augmented-State-Observer-Based Diagnostics of Open-Circuit and Sensor Faults in DFIG Wind Turbines

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.

Sensitivity analysis by differential importance measure for unsupervised fault diagnostics

Fault diagnostic approaches based on supervised classifiers are difficult to apply to safety-critical or new design systems, because they require the availability of labelled data collected when the systems operate abnormally, which is a rare situation. To address this challenge, we develop a novel unsupervised method for fault diagnostics based on a fault detection module and on sensitivity analysis . Specifically, the Differential Importance Measure (DIM) is originally used to quantify how much a signal is, or a set of signals are, responsible for the variation of the system health state. The proposed method is tested on simulated data from a wind turbine and on real data from a gas turbine. The advantages of the proposed fault diagnostic method are: (1) it can be developed using only normal condition data (2) it allows identifying the component responsible for the abnormality by quantifying the contribution of groups of signals to the variation of the system health state; (3) it is capable of distinguishing the abnormalities caused by changes in external conditions from those caused by components malfunctions; (4) it can be used in combination with any fault detection technique.

A Bayesian deep learning framework for reliable fault diagnosis in wind turbine gearboxes under various operating conditions

Vibration-based fault diagnostics combined with deep learning approaches has promising applications in detecting and diagnosing faults in wind turbine gearboxes. Specifically when time series vibration data is transformed to a 2-dimensional cyclic spectral coherence maps, the accuracy of deep neural networks in classifying faults increases. Nevertheless, standard deep learning techniques are vulnerable to inaccurate predictions when tested with new data originating from unseen faults or unusual operating conditions. To address some of these shortcomings in the context of wind turbine gearboxes, this paper investigates fault diagnostics using Bayesian convolutional neural network which provide accurate results with uncertainty bounds reducing wrong overconfident classifications. The performance of Bayesian and standard neural networks is compared using a simulation-based dataset of acceleration signals generated from a multibody dynamic model of a 5 MW wind turbine. The framework proposed in this paper has relevance to fault detection and diagnosis in other rotating machinery applications.

Hierarchical dynamic aggregation graph modeling of vibration for semi-supervised fault diagnosis of power transformer

Power equipment fault diagnostics is a critical aspect of ensuring the stability of the power grid system. However, it presents substantial challenges in obtaining labeled data that is spatial-temporal, multi-scale, and multi-domain, low noise for effective fault analysis and diagnosis. To address this issue, we propose a novel approach called hierarchical dynamic aggregation graph (HDAG) modeling for self-supervised fault diagnosis of power transformers using vibration data. Firstly, HDAG focuses on modeling the spatial and temporal correlations within the fault vectors, before converting them into time–frequency images for visualization. Secondly, our proposed fault diagnosis approach, (comprising the ST-sparse swin-transformer and multi-domain transformer fusion module), is integrated into the methodology. The ST-sparse swin transformer incorporates soft threshold modules, enabling the retention of relevant information while discarding irrelevant information. The multi-domain transformer fusion is designed to utilize the intra-domain and inter-domain signal characteristics to achieve a comprehensive feature representation. Finally, we present case studies based on experimental data that demonstrate the feasibility and effectiveness of our approach. Comparative evaluations against eight state-of-the-art techniques validate the improved information representation and diagnostic capabilities of our recommended strategy.