Condition Monitoring Information Research Articles

Joint learning strategy of multi-scale multi-task convolutional neural network for aero-engine prognosis

Remaining useful life (RUL) prediction and health status (HS) assessment are two key tasks in aero-engine prognostics and health management (PHM) system. However, existing deep learning-based prognostic models perform RUL prediction and HS assessment tasks separately, without considering the correlation between these two tasks. Secondly, traditional deep learning can only extract single-scale features, which limits the ability to extract complex degradation features from high-dimensional condition monitoring data. Therefore, this work proposes a multi-scale and multi-task convolutional neural network for joint learning of aero-engine RUL prediction and HS assessment. Firstly, multi-sensor data with multiple cycles are converted into image samples to integrate more condition monitoring information that is beneficial to prognosis. Then, the multi-scale feature fusion block is designed as the shared network for multi-task, utilizing convolutional layers with filters of different sizes to enhance the ability to extract complex degradation features from high-dimensional condition monitoring data. And a multi-layer concatenation block is constructed to integrate multi-scale features at different levels to fully utilize the important information at different levels. On this basis, a multi-task joint learning block is constructed and a joint loss function is developed for joint learning of RUL prediction and HS assessment. Finally, experiments on two engine degradation datasets, CMAPSS and N-CMAPSS, demonstrate that the proposed network has excellent RUL prediction and HS assessment performance, and outperforms other state-of-the-art methods.

Variational multi-harmonic duality mode pursuit method for extracting repetitive transient components from vibration signals

The repetitive transient component (RTC) in vibration signals is a typical symptom reflecting the localised defects of rotating machinery, which can be mathematically defined by feature parameters, such as the inter-arrival interval, maximum amplitude, resonance frequency, and attenuation ratio. In particular, the inter-arrival interval and maximum amplitude reveal the location and severity of failure and provide critical information for condition monitoring and fault diagnosis. However, there are some shortcomings of applicability, anti-noise, and fidelity, which are unfavourable for pursuing fault-induced RTC and enabling reliable recognition of key feature parameters. Hence, a novel method called variational multi-harmonic duality mode pursuit (VMHDMP) is proposed to extract the RTC accurately. A multi-harmonic duality mode function (MHDMF) that comprehensively depicts the critical feature parameters of the RTC was first defined in the frequency domain. Based on the defined MHDMF, the pursuit of RTC is formulated as a frequency-domain variational optimisation problem with a local compact time duration constraint, which optimally searches for the mode with an explicit MHDMF from the complex-valued frequency signal. A strategy involving the mutual conversion of real- and complex-valued signals was incorporated into the solution algorithm to satisfy the decomposability of complex-valued frequency signals. Finally, an inverse Fourier transform is performed on the analytical form of the extracted MHDMF to obtain a faithful RTC. Essentially, VMHDMP provides an optimal ‘transient-pass’ inverse filter bank with compact passage only near the occurrence moment of each transient in RTC. Thus, the vital parameters closely related to the faults are fully retained, whereas the frequency-overlap noise and large-amplitude interference occurring at different moments can be effectively eliminated. Simulated and experimental examples were used to test the effectiveness of VMHDMP.

A Few-Shot Learning Based Fault Diagnosis Model Using Sensors Data from Industrial Machineries

Efficient maintenance in the face of complex and interconnected industrial equipment is crucial for corporate competitiveness. Traditional reactive approaches often prove inadequate, necessitating a shift towards proactive strategies. This study addresses the challenges of data scarcity and timely defect identification by providing practical guidance for selecting optimal solutions for various equipment malfunction scenarios. Utilizing three datasets—Machine Sound to Machine Condition Monitoring and Intelligent Information (MIMII), Case Western Reserve University (CWRU), and Machinery Failure Prevention Technology (MFPT)—the study employs the Short-Time Fourier Transform (STFT) as a preprocessing method to enhance feature extraction. To determine the best preprocessing technique, Gammatone Transformation, and raw data are also considered. The research optimizes performance and training efficiency by adjusting hyperparameters, minimizing overfitting, and using the KERAS Early Halting API within resource constraints. To address data scarcity, which is one of the major obstacles to detecting faults in the industrial environment, Few-shot learning (FSL) is employed. Various architectures, including ConvNeXt Base, Large MobileNetV3, ResNet-18, and ResNet-50, are incorporated within a prototypical network-based few-shot learning model. MobileNet’s lower parameter count, high accuracy, efficiency, and portability make it the ideal choice for this application. By combining few-shot learning, MobileNet architecture, and STFT preprocessing, this study proposes a practical and data-efficient fault diagnosis method. The model demonstrates adaptability across datasets, offering valuable insights for enhancing industrial fault detection and preventive maintenance procedures.

Deep Learning-Based Cow Tail Detection and Tracking for Precision Livestock Farming

Cow tail detection and tracking in videos provide valuable information for individual identification, calving process, behavior analysis, and body condition monitoring. Although deep learning-based detection methods have demonstrated good performance, many of them have high complexity and still require an improvement in video object detection by reducing the computation time and false positives. To fully exploit the interframe information and achieve a real-time online detection, the optimized cow tail detection and tracking method is proposed based on an improved single shot multibox detector (SSD) and Kalman filter. Here, our improved SSD-integrated DenseNet and Inception-v4 reduce detection information loss and network parameters. Then, an improved window function-based Kalman filter and Hungarian are adopted to remove error detections and enhance the cow tail tracking accuracy. Experiments on our acquired rear-view videos show that the proposed approach achieved fast tail detection with an accuracy of 96.97%, a speed of 96 fps, a smaller model size of 25 MB, and higher position accuracy with an average 6.45 pixels deviation. The proposed approach outperformed the region-based R-CNN models and other tracking methods (e.g., particle filter), which provides a new solution to automatic cow detection and tracking in smart livestock farming.

Condition-based multi-component maintenance decision support under degradation uncertainties

Multi-component maintenance optimization is a well-studied area for age-based failure models but in contrast, incorporation of condition-based maintenance (CBM) is still an open area of research. Taking advantage of condition monitoring information for updating components’ health conditions demands a dynamic short-term approach when grouping multiple activities subject to CBM policy. Degradation models are commonly utilized in CBM for predicting the future condition of a given component to decide appropriate maintenance actions where inherent uncertainties exist in the degradation processes. There are a limited number of works in literature that account for degradation uncertainties where maintenance cost is a function of such uncertainty. This paper aims to develop a maintenance decision support for a multi-component system by incorporating CBM while considering the degradation uncertainties. In this paper, a two-stage stochastic programming is proposed to address such an issue and the problem is formulated for situations where maintenance opportunities are limited due to practical constraints (e.g., remote offshore maintenance operations of wind farms, unmanned platforms in oil and gas industries, etc.). The concept of marginal cost is used in developing the equation of optimality. This is a combinatorial problem and becomes intractable when the number of components is large therefore a heuristic is proposed to reduce the problem size which reduces the required computational time substantially. It is shown that significant cost savings are possible, especially, when the downtime cost and common setup cost are significant. A numerical example is provided with a system of six components achieving above 10%\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$10\\%$$\\end{document} cost reduction when the degradation uncertainties are taken into account.

Joint optimization of preventive and condition-based maintenance for offshore wind farms

High costs of maintenance and lost production during downtime are a challenge to the offshore wind industry, and there is a great potential to improve cost efficiency by improved maintenance and control strategies utilizing condition monitoring information. As wind farms get older, there is also an increased need to find ways of extending the lifetime of wind turbines allowing continued operation. This may be obtained by de-rating strategies, meaning adjustments of the power production to reduce the fatigue loads on the turbines. This subsequently means wind farm operators are faced with a trade-off between maximizing power production while limiting the degradation of the turbines. To investigate the best trade-off, this paper presents an optimization framework that considers component condition and planned power production to find the best times to perform predetermined preventive and condition-based maintenance on an offshore wind farm. To solve the scheduling problem, it is formulated as a constrained integer linear program, maximizing the net income for the planning horizon. The proposed method considers logistic restrictions, wind and electricity price forecasts, control strategies, component condition and probability of failure. Moreover, the method uses a short time horizon (days) to utilise weather forecasts and a long time horizon (weeks) to better capture the impact of deteriorating condition. The model is presented in a general framework for accounting for component condition in offshore wind farm operation and maintenance. It is illustrated for a specific potential application, considering condition monitoring of main bearings and corrosion of structural elements as examples.

The Optimization of Parallel Resonance Circuit for Wear Debris Detection by Adjusting Capacitance

Wear debris in lubrication oil provides important information for marine engine condition monitoring and faults diagnosis. Inductive sensors have been widely used to detect wear debris in lubrication oil. To improve the sensitivity, the inductive coil is always connected with a capacitor in parallel to form parallel LC resonance-sensing circuit. A previous study optimized the parallel resonance circuit by adjusting the excitation frequency. However, multiple parameters (namely, excitation signal, signal detection circuits, and signal-processing program, etc.) need to be adjusted accordingly for a series of the testing frequencies. To simplify the optimization, we propose a method based on adjusting the parallel capacitance in this work. The impedance (inductance and internal resistance) of the sensing coil and its variation induced by particles are first measured, which are the necessary parameters for establishing the function relationship between the parallel capacitance and the relative impedance variation. With the function relationship, the relative impedance variation is calculated directly, and the optimal capacitance is located at the highest absolute value of it. The experimental results for the ferrous and nonferrous particles match the calculation results well. Interestingly, the optimal capacitance for the nonferrous particle was larger than that of the ferrous particle. We speculate that the difference is generated due to the increased resistance induced by the eddy current effect.

Quasi-Distributed Temperature Detection of Press-Pack IGBT Power Module Using FBG Sensing

In situ temperature near the chips in a multiple-chip power semiconductor module is important information for thermal design validation, condition monitoring, and overheat protection. Due to the compact packaging structure and clamping force, online detection of the temperature is challenging for a press-pack module. This study attempts to achieve measurement by integrating optical fiber Bragg grating (FBG) sensors into the device. Different integration schemes are examined using experiments and theoretical analysis to show the effects on the transient performance. It is shown that FBG sensors with plate housing can effectively indicate the transient temperature ripple in normal converter operation, owing to the fast dynamic response of a small size sensor. The package integration has only minimal effects on the original distributions of electrical, thermal, and mechanical quantities within the power module.

Risk Control of Mission-Critical Systems: Abort Decision-Makings Integrating Health and Age Conditions

The mission abort is an effective action to reduce the risk of casualties and enhance the survivability of mission-oriented systems, such as aircrafts, submarines, and unmanned aerial vehicles (UAV). A key target of operators in real mission environment is to strive for balance between the success possibility of a mission and system survivability (SS), via elaborate mission abort plans. In this article, we design mission abort policies based on two crucial information: 1) degradation degree of monitored health features, and 2) system age. Accordingly, the operators may abort the mission if the degradation attains a preset control limit at early system ages, or continue the task otherwise. We carry out loss analysis to determine the optimal abort action, by linking MSP and SS. Structural insights regarding the optimality of degradation control limit as well as age threshold are explored. For a comparative purpose, the performances of some heuristic abort policies are analytically evaluated. We make contribution by scheduling mission abort plans harnessing both condition monitoring and age information, which promotes the timeliness and robustness of risk control. A case study on inertial navigation systems of UAV executing line inspection missions is used to illustrate the applicability and superiority of the proposed abort policies.

Microservices-based cloud-edge collaborative condition monitoring platform for smart manufacturing systems

In the context of the Industrial Internet of things (IIoT), large-scale IIoT data is generated, which can be effectively mined to provide valuable information for condition monitoring (CM). However, traditional CM methods cannot meet unprecedented challenges concerning large-scale IIoT data transmission, storage and analysis. Therefore, manufacturers have begun to shift from the traditional manufacturing paradigm to smart manufacturing, which integrates the encapsulated manufacturing services and the enabling cloud-edge computing technology to handle large-scale IIoT data. To enhance the agility, scalability and portability of traditional manufacturing services, a microservices-based cloud-edge collaborative CM platform for smart manufacturing systems is proposed. First, leveraging the microservices management system, the lightweight edge and cloud services are constructed from the microservices level, which enables flexible deployment and upgrade of services. Next, the proposed platform architecture effectively integrates the computing and storage capabilities of the cloud layer and the real-time nature of the edge layer, where the cloud-edge collaborative mechanism is introduced to achieve real-time diagnosis and enhance prognosis accuracy. Finally, based on the proposed system, the diagnosis and prognosis tasks are implemented on a manufacturing line, and the results show that the diagnostic accuracy is 90% and the prediction error is 50%.

Vision-based fringe projection measurement system for radial vibration monitoring of rotating shafts

Shaft radial vibration signals of rotary machines are crucial dynamic information for condition monitoring and fault diagnosis. A novel vision-based fringe projection vibration measurement system is proposed to monitor the radial displacements of a rotating shaft. The fringe pattern was projected on the surface of the shaft and the deformed fringe was acquired by a high-speed camera. A novel fringe processing algorithm was proposed to extract the radial displacement with good accuracy and efficiency. The vertical displacement could be obtained by locating the symmetric center of the recorded fringe. Meanwhile, the horizontal displacement could be obtained from the shift of the fringe at the symmetric center. The influence factors affecting measurement performance were analyzed. Simulations and comparative experiments were carried out for the verification of the feasibility and reliability of the proposed system. Comparing with the multiple eddy current sensors based system, the proposed method can realize multi-dimensional vibration measurement using one camera. The merits of simple in system, easy in operation, high accuracy and effectiveness in measurement make it attractive for dynamic characterization, condition monitoring and faults diagnosis of rotating machineries in actual applications.

Sensor-Based Prototype of a Smart Assistant for Visually Impaired People-Preliminary Results.

People with visual impairment are the second largest affected category with limited access to assistive products. A complete, portable, and affordable smart assistant for helping visually impaired people to navigate indoors, outdoors, and interact with the environment is presented in this paper. The prototype of the smart assistant consists of a smart cane and a central unit; communication between user and the assistant is carried out through voice messages, making the system suitable for any user, regardless of their IT skills. The assistant is equipped with GPS, electronic compass, Wi-Fi, ultrasonic sensors, an optical sensor, and an RFID reader, to help the user navigate safely. Navigation functionalities work offline, which is especially important in areas where Internet coverage is weak or missing altogether. Physical condition monitoring, medication, shopping, and weather information, facilitate the interaction between the user and the environment, supporting daily activities. The proposed system uses different components for navigation, provides independent navigation systems for indoors and outdoors, both day and night, regardless of weather conditions. Preliminary tests provide encouraging results, indicating that the prototype has the potential to help visually impaired people to achieve a high level of independence in daily activities.

Multimodal image registration for power equipment using clifford algebraic geometric invariance

Imaging inspection and video monitoring are important means for power equipment condition monitoring and diagnosis. However, single model image cannot supply sufficient information for power equipment condition monitoring, such as infrared images and visible images. The image registration can bridge this gap. This paper develops an image registration method fused by infrared and visible images, which constructs the principal direction of corner points using neighborhood contour point geometric invariance under Clifford algebraic space. Simultaneously, we also design a law for fitting optimal spatial information in the main direction of corner points, where the Clifford rotation operator is utilized to realize the spatial transformation of the image to be registered. The experiment reveals that the average registration error is less than 3 pixels by this developed method, reaching sub-pixel registration, subsequently obtaining satisfactory fusion effect of power equipment. This algorithm enriches the multimodal image registration scheme and meets the requirements of monitoring power equipment.

A Data-Driven Approach for Assessing Aero-Engine Health Status

Aero-engine health assessment is an important issue in the prognostics and health management (PHM) field. To improve the accuracy of health assessment, a novel data-driven approach is proposed in this paper. Unlike the model-based methods, the proposed data-driven method does not require to know the physical nature of degradation mechanism. Also, it is different from traditional data-driven methods and does not require to construct the synthesized health index or pre-set the health thresholds. In this work, the proposed method consists of feature selection, health state division and health state evaluation. Firstly, Fisher's discriminant ratio (FDR) is proposed to evaluate the sensor-based condition monitoring information, so as to eliminate the redundant information while retaining dominant information. Secondly, to solve the problem of unlabeled observation data, fuzzy c-means (FCM) clustering algorithm is presented to define the aero-engine health states. Finally, a bidirectional long short term memory (Bi-LSTM) network, which can capture the bidirectional long-range dependencies of features, is introduced to construct the health state evaluation model. Experimental results on aero-engine data set from NASA indicate that the proposed method can effectively improve the accuracy of health assessment, which can inform maintenance-related decisions.

The Deterministic Nature of Sensor-Based Information for Condition Monitoring of the Cutting Process

Condition monitoring of the cutting process is a core function of autonomous machining and its success strongly relies on sensed data. Despite the enormous amount of research conducted so far into condition monitoring of the cutting process, there are still limitations given the complexity underlining tool wear; hence, a clearer understanding of sensed data and its dynamical behavior is fundamental to sustain the development of more robust condition monitoring systems. The dependence of these systems on acquired data is critical and determines the success of such systems. In this study, data is acquired from an experimental setup using some of the commonly used sensors for condition monitoring, reproducing realistic cutting operations, and then analyzed upon their deterministic nature using different techniques, such as the Lyapunov exponent, mutual information, attractor dimension, and recurrence plots. The overall results demonstrate the existence of low dimensional chaos in both new and worn tools, defining a deterministic nature of cutting dynamics and, hence, broadening the available approaches to tool wear monitoring based on the theory of chaos. In addition, recurrence plots depict a clear relationship to tool condition and may be quantified considering a two-dimensional structural measure, such as the semivariance. This exploratory study unveils the potential of non-linear dynamics indicators in validating information strength potentiating other uses and applications.

An optimal opportunistic maintenance policy for a two-unit series system with general repair using proportional hazards models

This paper proposes an opportunistic maintenance policy for maintaining mechanical systems consisting of two series units with economic dependency. Condition monitoring and age information is available for unit 1 and unit 2, respectively. A gamma process describes the stochastic deterioration of unit 1, and condition monitoring data is utilized to estimate its hazard rate using the proportional hazards model (PHM). A matrix-based algorithm is employed to derive the conditional reliability of a two-unit series system with a continuous-state deterioration process. The policy proposes preventive actions on scheduled intervals, based on lifetime and PHM parameters, to enhance system reliability and mitigate unexpected downtimes. Corrective actions rectify the failed unit, while the other unit can be maintained opportunistically. The problem is formulated as a semi-Markov decision process (SMDP) with an objective of minimizing the long-run expected average cost over an infinite time horizon. This paper relaxes a common assumption that repair brings the units to the as-new condition, and provides solutions for the maintenance scheduling of two-unit series systems, where repair is also allowed.

Multi-Information Fusion-Based Hierarchical Power Generation-Side Protection System

With renewable power sources and new topology structures being widely introduced into the power system, the current local information-based power generation-side protection cannot fully guarantee the protection performance and the safety coordination with the power grid. This paper proposes an improved hierarchical protection system on the power generation-side. The proposed system takes advantage of the fusion of multi-information provided by the system, station and local layers. The system layer provides the information such as system voltage control and power regulation demand, so that the generation-side protection and control system can adapt to the system operation mode and power regulation demand. The station layer realizes the coordination of the protection principle and action strategy among the related units through information interaction with the power grid, the automatic control system and the local layer protection. The local layer introduces the condition monitoring information and more abundant protection information to enhance the protection performance and master the generation units’ safety condition. To illustrate the hierarchical protection system construction method and actual application mode, the multi-information fusion-based comprehensive local layer protection method and the multi-generators information fusion-based hierarchical protection method are taken as examples. Case analysis shows that the proposed methods can reflect more slight internal fault forms and can adaptively determine the protection action characteristic and tripping strategy according to the system operating conditions and other generators’ fault conditions, which effectively improve the protection sensitivity and coordination capability. To provide reference and inspiration for follow-up research, the hierarchical protection system construction mode, communication technology and research approaches of new protection methods are further pointed out.

Entity Extraction of Electrical Equipment Malfunction Text by a Hybrid Natural Language Processing Algorithm

Many electrical equipment malfunction text messages are collected during power system operation and maintenance procedures. These texts usually contain crucial information for maintenance and condition monitoring. Because these power system malfunction texts are characterized by multidomain vocabularies, complex-syntactic structures, and long sentences, it is challenging to for automated systems to capture their semantic meaning and essential information. To address this issue, we propose a hybrid natural language processing (hybrid-NLP) algorithm to extract entities that represent electrical equipment. This algorithm is composed of a dictionary-based method, a language technology platform (LTP) tool, and the bidirectional encoder representations from a transformers-conditional random field (BERT-CRF) model. Significantly, the softmax output layer of the bidirectional encoder representations from the transformers (BERT) model is replaced by the conditional random field (CRF) algorithm to strengthen the contextual relationships between words and thus solve the local optimization of the preferred word label. The effectiveness of the proposed hybrid-NLP method is verified on a realistic dataset. Moreover, a statistical analysis is conducted to provide a reference for the operation and maintenance of power systems.

APPLICATION OF ULTRA-HIGH FREQUENCY ANTENNA FOR PARTIAL DISCHARGE DETECTION IN OIL-FILLED POWER TRANSFORMER: A REVIEW

Partial discharge (PD) is a phenomenon that may lead to dielectric breakdown and can provide important information for condition monitoring on electrical power equipment, in particular transformer. One of the methods is the detection of the electromagnetic (EM) wave signal emitted by PD. Although the frequency spectrum in E is very wide, this paper discusses the detection of EM only at ultra-high frequency (UHF). One of the detectors that can be used to detect EM is the antenna. There are a lot of antenna designs that have been proposed to detect the signal. The designs can be generally divided into two: PCB-based design and physical antenna design. An example of the latter is monopole. Some of the proposed antennas were left at the design stage while others went to be applied in actual PD experimentation. Discussion on the capabilities of these antennas can lead to the selection of a suitable antenna.

Probabilistic cost‐benefit analysis‐based spare transformer strategy incorporating condition monitoring information

Determining an economical and reliable spare transformer strategy is an essential and challenging task for power utility companies. This study proposes a probabilistic cost-benefit analysis-based spare transformer strategy incorporating condition monitoring information. Not only the actual wearing condition of each transformer, but also the impacts of a transformer ageing process on spare transformer strategy are considered by the proposed method. First, the actual wearing condition and functional age of each transformer are estimated based on condition monitoring information. Secondly, the ageing process and ageing failure rate of each transformer are modelled by combining a Weibull distribution and its functional age. Thirdly, based on probabilistic cost-benefit analysis, the outage costs, as well as economic benefits under different inventory sizes, are estimated. The transformer's ageing process is modelled and considered in the determination of the optimal number and the acquisition timing of spare transformers. Finally, a transformer substation group and its condition monitoring information, and the real spare transformer strategy collected from a regional power grid in China were used to demonstrate an application of the proposed method. The results indicate that the spare transformer strategy determined by the proposed method not only enables the power utilities to avoid economic losses but also guarantees a high cost-benefit ratio of spare transformer strategy.