Search

Time series anomaly detection (TSAD) is an essential problem faced in several fields, e.g. fault detection, fraud detection, and intrusion detection, etc. Although TSAD is a crucial problem in anomaly detection, few solutions in anomaly detection are suitable for it at present. Recently, some researchers use GAN-based methods such as TAnoGAN and TadGAN to solve TSAD problem. However, problems such as model collapse, low generalization capability and poor accuracy still exist. In this paper, we proposed a Dilated Convolutional Transformer-based GAN (DCT-GAN) to enhance accuracy and improve generalization capability of the model. Specifically, DCT-GAN utilize several generators and a single discriminator to alleviate the mode collapse problem. Each generator consists of a dilated convolutional neural network and a Transformer block to obtain fine-grained and coarse-grained information of the time series, which is a useful component to improve generalization capability. We also use weight-based mechanism to balance these generators. Experiments verify the effectiveness of our method and each part of DCT-GAN.

Fault diagnostics involving the Internet of Things (IoT) sensors and edge devices is a challenging task due to their limited energy and computational capabilities. Another challenge concerning IoT sensors or devices is the incursion of noise when used in an industrial environment. The noisy samples affect the decision support system that could lead to financial and operational losses. This paper proposes a Noisy Encoder using Artificial Intelligence of Things (NEAT) architecture for fault diagnosis in IoT edge devices. NEAT combines autoencoders and Inception module to co-train the clean and noisy samples for solving the said problem. Experimental results on benchmark datasets reveal that the NEAT architecture is noise resilient in comparison to the existing works. Furthermore, we also show that the NEAT architecture has lightweight characteristics as it yields a lower number of parameters, weight storage, training, and testing times that support its real-life applicability in an Industrial IoT environment.

In this paper, a fault detection method based on chaos is proposed for doubly-fed induction generator (DFIG). Firstly, the nonlinear differential equation of the DFIG is established by the stator flux-oriented vector control method, and then the motion state of the DFIG is proved by the 0-1 test algorithm. Secondly, the motion state of the DFIG is analyzed under normal and fault conditions. Due to some failures that may occur in the normal operation of the DFIG, the motion state of the DFIG system may enter chaos from the cycle, which is mainly manifested by the violent oscillation of the rotor current and speed. At this time, the system is in an extremely unstable operating condition. Thereby, a new sliding mode control rate is mentioned to suppress chaotic behavior. Its stability is analyzed by constructing the Lyapunov function to ensure that the system state can quickly and stably converge to the given value. After that, the simulation model of the system is built. The simulation results show that sliding mode control can make the DFIG get out of the chaotic state quickly. And, the sliding mode control has the characteristics of fast fault elimination and high control accuracy. In addition, the traditional sliding mode approach rate is compared with the proposed new sliding mode approach rate. It proves that the proposed new approach rate can achieve better control effect.

This paper highlighted on a new improvement of the salp swarm algorithm (SSA) that is taken from the behavior of salp fishes. This shows a promising result, however it suffered from some drawbacks that limited its ability. The proposed improvement of SSA is called SSA 2. In this paper the design system where the ZigBee protocol received temperature, humidity and gas from the sensors. The base station collects received data and sends it to platform in order to process these data via SSA2. This system enables the ability to monitor and track any fault that may be occurring. A new term has been developed to make the algorithm more targeted in finding the optimal answer. Moreover, the calculation results and simulation results in Matlab shows great enhancement in the accuracy and precision of the results at very fast speed for convergence compared with the original salp swarm algorithm (SSA).

In practical mechanical fault detection and diagnosis, it is difficult and expensive to collect enough large-scale supervised data to train deep networks. Transfer learning can reuse the knowledge obtained from the source task to improve the performance of the target task, which performs well on small data and reduces the demand for high computation power. However, the detection performance is significantly reduced by the direct transfer due to the domain difference. Domain adaptation (DA) can transfer the distribution information from the source domain to the target domain and solve a series of problems caused by the distribution difference of data. In this survey, we review various current DA strategies combined with deep learning (DL) and analyze the principles, advantages, and disadvantages of each method. We also summarize the application of DA combined with DL in the field of fault diagnosis. This paper provides a summary of the research results and proposes future work based on analysis of the key technologies.

Accurate and timely fault diagnosis of the rotating rectifier for a wound-rotor synchronous starter/generator is of great importance to improve system reliability. In the generation mode of the system, there is one idle stator winding of the main exciter. The terminal voltage characteristic of the idle stator winding is analyzed, and a novel fault diagnosis method for the rotating rectifier in the generation mode of the system is proposed. In this method, two feature voltages, namely, the maximum peak voltage and any other peak voltage of the idle stator winding terminal voltage are extracted through voltage sequence reconstruction to calculate the feature ratio for fault detection. The corresponding rotor position obtained by an inherent position sensor where the maximum peak voltage occurs is collected to determine the faulty diode location. The experimental results verify the feasibility and effectiveness of the proposed fault diagnosis method.

Data-driven fault detection and isolation (FDI) depends on complete, comprehensive, and accurate fault information. Optimal test selection can substantially improve information achievement for FDI and reduce the detecting cost and the maintenance cost of the engineering systems. Considerable efforts have been worked to model the test selection problem (TSP), but few of them considered the impact of the measurement uncertainty and the fault occurrence. In this article, a conditional joint distribution (CJD)-based test selection method is proposed to construct an accurate TSP model. In addition, we propose a deep copula function which can describe the dependency among the tests. Afterward, an improved discrete binary particle swarm optimization (IBPSO) algorithm is proposed to deal with TSP. Then, application to an electrical circuit is used to illustrate the efficiency of the proposed method over two available methods: 1) joint distribution-based IBPSO and 2) Bernoulli distribution-based IBPSO.

This article proposes an integrated approach of model-based and data-driven gap metric fault detection and isolation in a stochastic framework. For actuator and sensor faults, an adaptive Kalman filter combining with the generalized likelihood ratio method is suggested. For component faults, especially incipient faults, the model-based scheme maybe not a good choice due to the existence of disturbances or noises. Hence, a novel data-driven gap metric strategy is presented. The design of the appropriate fault cluster center model and radius via the gap metric technique is put forward to enhance the isolability of the incipient faults. Numerical simulation results are given to demonstrate the effectiveness of the proposed fault detection and isolation algorithm.

Fault detection and degradation modeling are two main concerns in condition-based maintenance (CBM). The initial machine degradation is called first predicting time (FPT) or incipient failure time. FPT is typically assumed prior information. FPT detection aims to provide such prior information for subsequent degradation modeling in CBM. Moreover, the majority of existing methodologies regard FPT detection and degradation modeling as two separate tasks. A generic framework for integration of the FPT detection with degradation modeling is proposed in this article via fusion of spectrum amplitudes in the frequency domain to realise FPT detection and degradation modeling in a unified manner. First, a generalised health index is constructed using the sum of weighted spectrum amplitudes. Second, two properties are proposed to describe FPT detection and degradation modeling. Third, these two properties and their constraints are mathematically formulated as a quadratic programming model to find optimal weights for the fusion of spectrum amplitudes automatically. Finally, three illustrative examples are used to demonstrate the superiority of the proposed methodology over some existing commonly used sparse measures and a machine learning method in the FPT detection and degradation modeling.

Aiming at the misdiagnosis problem of the inter-turn short-circuit (ITSC) fault detection method caused by the inverter nonlinearity and the current measurement errors, a d-axis residual voltage-based fault diagnostic strategy is proposed for sensorless control of the surface permanent magnet synchronous motor (SPMSM). The model of the 3-phase wye-connected SPMSM with ITSC fault is established, and the expression of the d-axis residual voltage caused by the ITSC fault, the inverter nonlinearity, the current measurement errors and the motor parameter errors is derived. The proposed fault diagnostic strategy can eliminate the current relative gain coefficient error and locate the ITSC fault by using the 8th and 2nd harmonics of the estimated d-axis residual voltage, respectively. The experimental results verify the effectiveness of the proposed fault diagnostic strategy.

Many methods for monitoring multivariate processes are built on principal component analysis (PCA), which, however, simply tells whether the process is faulty or not. In fact, there is still room for the improvement of the early detection performance by exploiting fully the information given by fault directions. To this end, this article proposes a novel directional PCA (diPCA) approach. First, by narrowing down faults to a specified direction or composite mutually orthogonal directions, diPCA can speed fault detection and facilitate accurate fault diagnosis. It also has good theoretical properties that guarantee concise control limits. Second, with appropriate fault directions, diPCA provides a unified framework for process monitoring and includes existing monitoring indices, such as Hotelling's T² and the squared prediction error (SPE), as special cases. Third, diPCA also naturally results in a new combined monitoring statistic, which is composed of both T² and SPE, and provides an optimal ratio of their combination. The Monte Carlo simulation results have demonstrated the power of the proposed monitoring and diagnostic methods stemming from diPCA. The proposed methods have also been implemented into the Tennessee Eastman process.

Wireless Body Area Networks (WBAN) can provide continuous monitoring of patients’ health. Such monitoring can be a decisive factor in health and death situations. Fault management in WBANs is a key reliability component to make it socially acceptable and to overcome pertained challenges such as unpredicted faults, massive data streaming, and detection accuracy. Failures in fault detection due to hardware, software, and network issues may put human lives at risk. This paper focuses on detecting and predicting faults in sensors in the context of a WBAN. A framework is proposed to manage AI-based prediction models and fault detection using thresholds where four Machine learning techniques: Artificial Neural Networks (ANN), Deep Neural Networks (DNN), Support Vector Machines (SVM), and Decision Trees (DT), are used. The framework also provides alarm notifications, prediction model deployment, version control, and sensing node profiling. As a proof of concept, a fault management prototype is implemented and validated. The prototype classifies faults, manages automation of sensing node profiling, training, and validation of new models. The obtained experimental results show an accuracy greater than 96% for detecting faults with an inferior false alarm rate.

With the aid of neural networks, this article develops two data-driven designs of fault detection (FD) for dynamic systems. The first neural network is constructed for generating residual signals in the so-called finite impulse response (FIR) filter-based form, and the second one is designed for recursively generating residual signals. By theoretical analysis, we show that two proposed neural networks via self-organizing learning can find their optimal architectures, respectively, corresponding to FIR filter and recursive observer for FD purposes. Additional contributions of this study lie in that we establish bridges that link model- and neural-network-based methods for detecting faults in dynamic systems. An experiment on a three-tank system is adopted to illustrate the effectiveness of two proposed neural network-aided FD algorithms.

This study focuses on dissipativity-based fault detection for multiple delayed uncertain switched Takagi-Sugeno fuzzy stochastic systems with intermittent faults and unmeasurable premise variables. Nonlinear dynamics, exogenous disturbances, and measurement noise are also considered. In contrast to the existing study works, there is a wider range of applications. An observer is explored to detect faults. A controller is studied to stabilize the considered system. A piecewise fuzzy Lyapunov function is collected to obtain delay-dependent sufficient conditions by means of linear matrix inequalities. The designed observer has less conservatism. In addition, the strict (Q, S,R)-e-dissipativity performance is achieved in the residual dynamic. Besides, the elaborate H∞ performance and the elaborate H performance are also acquired. Finally, the availability of the method in this study is verified through two simulation examples.

During the manufacturing process of three-dimensions integrated circuits (3-D ICs), many defects may occur in through-silicon vias (TSVs). These defects affect integrity of the signal passing through TSVs, therefore detecting these defects in early production stage is crucial. This paper proposed a built-in-self-test (BIST) method with simple test structure and easily-implemented fault detection process. Due to the fact that the defects in TSVs influence the charge/discharge speed of TSVs’ equivalent capacitance, a weak current source is designed as the central part of the proposed test structure, which can measure the TSVs’ charge/discharge speed and quantify it using digital values. The effectiveness and robustness of the proposed method is validated by HSPICE simulation. The minimum measurable fault, test time and test circuit area are also assessed in this paper. The simulation results show that the method this paper proposed can detect weaker faults, compared with the traditional test methods.

PUBLISHED ON IEEE/ASME TRANSACTIONS ON MECHATRONICS, DOI: 10.1109/TMECH.2021.3100150. Ideally, accurate sensor measurements are needed to achieve a good performance in the closed-loop control of mechatronic systems. As a consequence, sensor faults will prevent the system from working correctly, unless a fault-tolerant control (FTC) architecture is adopted. As model-based FTC algorithms for nonlinear systems are often challenging to design, this paper focuses on a new method for FTC in the presence of sensor faults, based on deep learning. The considered approach replaces the phases of fault detection and isolation and controller design with a single recurrent neural network, which has the value of past sensor measurements in a given time window as input, and the current values of the control variables as output. This end-to-end deep FTC method is applied to a mechatronic system composed of a spherical inverted pendulum, whose configuration is changed via reaction wheels, in turn actuated by electric motors. The simulation and experimental results show that the proposed method can handle abrupt faults occurring in link position/velocity sensors. The provided supplementary material includes a video of real-world experiments and the software source code., 11 pages, 7 figures

Online fault detection of insulators is a necessary requirement for the development of a smart grid, which directly affects the safety and reliability of power system operations. Breakage is an important factor causing insulator abnormalities, and a large number of image samples are often required for training to complete defect identification with the help of AI deep learning. In this paper, we propose a convolutional neural network ensemble implemented using small samples, data augmentation and transfer learning for insulator breakage detection. We used data augmentation to expand the small sample of 200 insulator images by 15 times to 3000 images, built a VGG16 transfer network, pre-trained the VGG16 network convolutional basis weights, and improved the training performance and efficiency of the model with freeze training and fine-tuning training. The training on 2700 images and testing results on 300 images show that the proposed method can achieve an accuracy of 0.9871 for insulator breakage recognition.

This article is concerned with the development of a new event-triggered parity space fault detection (FD) scheme. A linear discrete-time system model with varying sampling periods is presented for handling the problem of event-triggered FD and a new parity relation is established. Based on this, an event-triggered residual generator is constructed and the generated residual is completely decoupled from event-triggered transmission error. The design of the parity matrix is formulated into an optimization problem and an optimal solution of the parity matrix is obtained by using singular value decomposition. The issue of residual evaluation is also considered in the event-triggering implementation. The novelties of this article are twofold. First, a new event-triggered parity relation is obtained and the parity space-based residual signal achieves complete decoupling with the event-triggered transmission error. Second, the calculation of the parity matrix is independent of event parameters. So the design of the parity space-based residual generator and event generator can be carried out independently. Finally, a simulation example is considered to demonstrate the effectiveness of the proposed method.

A fuzzy singular perturbation model with additive faults on control inputs and external disturbances, a fuzzy $H_\infty$ fault detector (FHFD), and a fuzzy $H_\infty$ fault-tolerant controller (FHFTC) are constructed for nonlinear singularly perturbed systems (NSPSs) with actuator faults and external disturbances, respectively. The sufficient conditions for FHFD and FHFTC existence are derived respectively, which are described by a set of $\sigma$ -independent linear matrix inequalities (LMIs), such that it can avoid the ill-conditioned problems caused by the tiny parameter $\sigma$ . An application to a CE150 helicopter with abrupt and intermittent faults is provided to verify the above approaches. The innovation of the paper lies in the following points: i) The FD and FTC for NSPSs are handled under the united model, which can avoid the incomplete decoupling problem caused by decomposing NSPSs as the slow and fast subsystems. ii) The estimate values for actuator faults are regarded as a state vector of the FHFD, and the output feedback control law is introduced into FHFD, which make the FHFD not only can accurately detect faults but also can stabilize the closed loop system. iii) A new method to transform the nonlinear matrix inequalities into the LMIs is proposed. iv) The presented FHFTC can avoid the problems of missing or inaccurate status information caused by the sensor faults because it adopts the state feedback of the FHFD.

This paper describes a novel short-circuit (SC) fault detection approach to protect the low-voltage dc microgrid (DCMG). The SC faults are the most common fault in the dc power system and can cause severe hazards if not isolated. Since a DCMG is a capacitor-dominated grid, the proposed scheme utilizes these filter capacitors current dynamics. An LV-DCMG system is considered for building the application to realize the proposed fault detection scheme. Active sources such as solar photovoltaic and battery are connected to the passive loads through a dc-dc converter, dc cable and solid-state relays. The proposed method considers an average of the capacitor current as a detection parameter. An iteration-based isolation configuration is also suggested for a zonal type of distributed network. Furthermore, a selection procedure of the thresholds is also presented in detail. The validity of the proposed approach is verified through both digital simulation studies and DSP- based experimental studies. The proposed approach is capable of detecting both low and high impedance faults.

Line-commutated converters are extensively used as the interface between ac grids and classic HVDC systems. At the inverter side, commutation failure of switches is one of the most common faults that can pose threats to the system operation. Practical and reliable study of such phenomena relies on accurate and efficient converter models for simulations. Recently, a parametric average-value model (PAVM) has been presented for acdc rectifiers, which considers the internal faults of the converter. In this paper, the PAVM methodology is extended to the dcac inverter systems, including the commutation failure of switches. The proposed PAVM also augments an automatic fault detection technique to determine the faulty switches. Using comprehensive simulation studies, the developed model is verified to accurately predict the commutation failure of switches and reconstruct the waveforms consistent with the detailed switching models of inverters while being computationally more efficient. The proposed PAVM is envisioned to be an efficient and accurate asset for simulation of HVDC systems and inevitable when faults of switches need to be considered.

In this article, fault diagnosis is of great significance for system health maintenance. For real applications, diagnosis accuracy suffers from unbalanced data patterns, where normal data are usually abundant than anomaly ones, leading to tremendous diagnosis obstacles. Therefore, it is challenging to use only normal data for fault diagnosis under this imbalanced condition. In addition, a single fault diagnosis model can only conduct one fault diagnosis task in most of cases. Accordingly, a one-class generative adversarial detection (OCGAD) framework based on semisupervised learning is proposed to learn one-class latent knowledge for dealing with multiple semisupervised fault diagnosis tasks, i.e., fault detection using only normal knowledge learning, novelty detection from unknown conditional data, and fault classification with unlabeled data. A bi-directional generative adversarial network (Bi-GAN) is first trained with only normal data. A one-class support vector machine is then established using features exacted by Bi-GAN from signals acquired from an attitude sensor for multifunctional fault detection. The presented OCGAD model is validated using an industrial robot with experiments of three fault detection tasks. The results demonstrate that the present model has good performance for dealing with multiple semisupervised diagnosis problems. info:eu-repo/semantics/publishedVersion

Dc collection and transmission is one of the main developing directions of the large-scale offshore wind power system. This work proposes a multi-functional dc collector to construct a low-cost, high reliability, and high flexibility all-dc offshore wind power system. By introducing the multi-functional dc collector to achieve energy collection and cascade boost, not only the system cost, size, and intermediate links of power conversion are reduced but also the coupling among wind turbines (WTs) is greatly weakened. Accordingly, the maximum power point tracking operation can be achieved independently for each WT unit. Furthermore, a constant voltage ratio control with first harmonic elimination iterative algorithm is proposed for dc collector to link the HVDC bus voltage and discretized dc voltage for triggering the energy consumption operation and suppressing the inductor current ripple, and the detailed control parameter design is provided. Meanwhile, the various operating modes are considered for multi-functional dc collector, including practical engineering configuration, fault isolation, dynamic cut-in/out, and energy consumption operation, which provides a practical solution for the construction of a long-distance and large-capacity offshore system. Finally, the comprehensive experimental results are given to illustrate the effectiveness of multi-functional dc collector.

Weather-related or other types of faults can negatively impact the continuity of electricity delivery to consumers and, lead to poorer reliability scores. 90% of customer outage-minutes are due to events which affect local distribution systems. The radial configuration is simple and has a low cost design; however, when faults occur, it often causes power outages beyond the location of the fault itself. It is beneficial for utilities to automatically isolate the fault and, thus, restore power to as many consumers as possible. The benefits of fault isolation and service restoration include: enhanced reliability and resiliency, reduced loss of revenue to the utility, and enhanced comfort and satisfaction for the utilitys consumers. This paper provides an algorithm to quickly isolate a fault and automatically restore power to as many nodes of the distribution system as possible by opening and closing switches already placed within the system. The expert model and centralized algorithm that are proposed in this paper guarantee an optimal solution and a short run-time regardless of the size of the distribution system. A mathematical proof of the optimal nature of the solution is provided. Also, the model is validated in practice, solving two different-sized systems.

Solid State Transformers (SSTs) are developing as highly efficient interfaces in renewable energy, transport, and energy storage systems. However, performance limitations, such as overvoltage sensitivity and fault handling capabilities, have slowed widespread adoption. Though SSTs are developing added capabilities for fault management, the required response speed and overdesign introduces added costs, particularly in protection of in ultra-low inductance systems, such as those sourced by energy storage systems (ESSs). With increased use of power electronics, the power distribution systems are speeding up having shorter fault propagation delays and higher fault currents. This creates a need for alternative approaches to MV system protection. This paper describes a Bidirectional Solid-State Circuit Breaker (BSSCB) based on a new SiC SuperCascode power switch, and a multi-layered transient absorption network. The paper studies transient heat transfer in the switch and provides a redefinition of the fuse curve as applied to the BSSCB suitable for digital control. The paper identifies critical fault issues, discusses the impact on critical design points of the SuperCascode for a BSSCB, and provides design calculations for a complete 10kV/300A/3X breaker, including the SuperCascode module. A scaled 6kV/10A/7X SuperCascode is fabricated and tested to demonstrate switch response of

In this article, an adaptive output feedback compensation scheme is proposed for a class of uncertain nonlinear systems with a set of redundant actuators and a sensor. The sensor output is the only information available for output feedback. Both the actuators and the sensor may suffer from failures during operation and, therefore, all the system states are not available. To estimate the unmeasured states, a modified high-gain observer is designed which can not only ensure the closed-loop system stability but also improve the tracking performance. In the adaptive controller design, to overcome the complexity due to the unknown actuator and sensor failures coupled with the nonlinear functions and unknown parameters, an upper bound estimation method is adopted, which greatly simplifies the design procedure. It is proved that with the proposed scheme, the actuator and sensor failures can be compensated simultaneously, the reconfiguration can be adaptively updated without fault detection and isolation, and the real-tracking error can converge to a residual set, that is, mainly proportional to the sensor gain variation and measurement error.

Modern industrial plants generally consist of multiple manufacturing units, and the local correlation within each unit can be used to effectively alleviate the effect of spurious correlation and meticulously reflect the operation status of the process system. Therefore, the local correlation, which is called spatial information here, should also be taken into consideration when developing the monitoring model. In this study, a cascaded monitoring network (MoniNet) method is proposed to develop the monitoring model with concurrent analytics of temporal and spatial information. By implementing convolutional operation to each variable, the temporal information that reveals dynamic correlation of process data and spatial information that reflects local characteristics within individual operation unit can be extracted simultaneously. For each convolutional feature, a submodel is developed and then all the submodels are integrated to generate a final monitoring model. Based on the developed model, the operation status of the newly collected sample can be identified by comparing the calculated statistics with their corresponding control limits. Similar to the convolutional neural network (CNN), the MoniNet can also expand its receptive field and capture deeper information by adding more convolutional layers. Besides, the filter selection and submodel development in MoniNet can be replaced to generalize the proposed network to many existing monitoring strategies. The performance of the proposed method is validated using two real industrial processes. The illustration results show that the proposed method can effectively detect process anomalies by concurrent analytics of temporal and spatial information.

Nowadays, how to select the kernel function and their parameters for ensuring high-performance indicators in fault diagnosis applications remains as two open research issues. This paper provides a comprehensive literature survey of kernel-preprocessing methods in condition monitoring tasks, with emphasis on the procedures for selecting their parameters. Accordingly, twenty kernel optimization criteria and sixteen kernel functions are analyzed. A kernel evaluation framework is further provided for helping in the selection and adjustment of kernel functions. The proposal is validated via a KPCA-based monitoring scheme and two well-known benchmark processes.

The fault detection (FD) problem for systems with both model uncertainty and external disturbance is investigated in this article. First, the mathematical models of systems with model uncertainty and disturbance, systems with additive faults, and systems with multiplicative faults are established with both left and right coprime factorization. Then, an observer-based FD scheme is proposed and the FD thresholds are derived for both open-loop and closed-loop manners. The necessary conditions on multiplicative FD are obtained and the fault detectability analyses are carried out with the aid of the gap metric technique. Finally, the effectiveness of the proposed method is illustrated by a case study on a cart dynamic system.

As industrial technology develops, industrial processes become increasingly large and complex, the traditional methods are difficult to extract features that can represent the condition of the whole process and the effect of fault on quality indicators. Therefore, a novel multi-block decouple convolutional neural network (multi-block DCN) algorithm is proposed. First, key process variables are selected, and process variables are grouped into multiple blocks for the following monitoring. Then, in each block, the proposed DCN constructs a regression model between key process variables and quality indicators, in which the regression model utilizes an improved convolutional neural network as feature extractor and a decoupling layer as feature regularizer. Afterward, the monitoring results of each block are integrated into a global monitoring index based on Bayesian theory. After fault detection, variable oblivion contribution plot is presented to locate faulty variables. Finally, an industrial case is used to demonstrate the effectiveness of multi-block DCN.

This paper is concerned with the finite frequency H-/H∞ memory fault detection filtering problem for discretetime Takagi-Sugeno (T-S) fuzzy affine systems with normbounded uncertainties. The objective is to design a piecewise affine memory filter by using system historical information such that the resulting closed-loop filtering error system is asymptotically stable with the prescribed finite frequency H-/H∞ performance. Based on the generalized Kalman-YakubovicPopov (GKYP) lemma combined with the celebrated S-procedure, new sufficient conditions for the fuzzy affine filtering error system to have the finite frequency H-/H∞ performance are given at first. By further using piecewise fuzzy quadratic Lyapunov functions (PFQLFs) and Projection lemma, the filtering analysis results for the filtering error system to be asymptotically stable with the prescribed finite frequency H-/H∞ performance are obtained. Then the filtering synthesis is carried out with the aid of matrix inequality convexification techniques, and the synthesis results are described in terms of linear matrix inequalities (LMIs). It is further shown that better filtering performance can be achieved by using more system historical information. Finally, simulation is provided to verify the effectiveness of the proposed approach.

Based on the interval type-2 fuzzy (IT2F) approach, this paper investigates the fault detection filter design problem for a class of nonhomogeneous higher-level Markov jump systems with uncertain transition probabilities. Considering that the mode information of the system cannot be obtained synchronously by the filter, the hidden Markov model can be seen as a detector to handle this asynchronous problem, and the parameter uncertainty can be processed by the IT2F approach with the lower and upper membership functions. Then, the asynchronous IT2F filter is designed to deal with the fault detection problem. Furthermore, the Gaussian transition probability density function is introduced to describe the uncertainty transition probabilities of the system and the filter. Based on Lyapunov theory, the existence of the designed asynchronous IT2F filter and the dissipativity of the filter error system can be well ensured. The simulation study on a quarter-car suspension system verifies that the designed asynchronous IT2F filter can detect faults without error alarms.

This paper proposes a neural network-based fault diagnosis scheme to address the problem of fault isolation and estimation for the Single-Gimbal Control Moment Gyroscopes (SGCMGs) of spacecraft in a periodic orbit. To this end, a disturbance observer based on neural network is developed for active anti-disturbance, so as to improve the accuracy of fault diagnosis. The periodic disturbance on orbit can be decoupled with fault by resorting to the fitting and memory ability of neural network. Subsequently, the fault diagnosis scheme is established based on the idea of information fusion. The data of spacecraft attitude and gimbals position are combined to implement fault isolation and estimation based on adaptive estimator and neural network. Then, an adaptive sliding mode controller incorporating the disturbance and fault estimation results is designed to achieve active fault-tolerant control. In addition, the paper gives the proof of the stability of the proposed schemes, and the simulation results show that the proposed scheme achieves better diagnosis and control results than compared algorithm.

Since the dc-side fault in the multi-terminal dc grid propagates extremely fast, the fault current limiter (FCL) is essential to achieve fault ride-through of the healthy network. Current-limiting inductors (CLIs) are currently the most commonly used FCLs in projects that have been built, are under construction, or are planned. The CLIs can be designed based on the system requirements, such as converter fault ride-through and dc circuit breaker (DCCB) interrupting capacity. Unfortunately, the installation of CLIs has negative impacts on the system transient response and operational stability, the fault isolation speed and energy dissipation of DCCBs. To solve these problems, an economical bridge-type multiport fault current limiter (BT-MFCL) is proposed in this paper, which can be installed at the dc bus with multiple dc lines connected and can cooperate with the existing DCCBs. The BT-MFCL is composed of a CLI on each port and a shared energy dissipation circuit (EDC). The fault current can be limited by the embedded CLIs in the BT-MFCL without a time delay, while the EDC can eliminate the negative influences of CLIs. Extensive simulations and experiments have verified the effectiveness of the BT-MFCL. Besides, the performance comparison of the BT-MFCL with existing solutions is elaborated.

Magnetic bearing is an important part of modern high-speed rotating machinery. The performance of power electronic converter highly determines the quality of magnetic bearing drive. In this paper, a fault-tolerant converter with series-winding topology is introduced. Compared with the traditional topology, the current stress of the phase leg is reduced while the utilization of DC voltage is guaranteed, and fault-tolerant operation can be realized in the case of open circuit fault of the switching device. This paper introduces the characteristics of the series-winding topology and the current control method. Open circuit fault detection methods and countermeasures are also analyzed. Experimental results show the effectiveness of the proposed topology. Fault-tolerant experiments under static suspension and rotation of the rotor demonstrate the performance of the topology. This kind of series-winding topology provides a new scheme with good performance and high reliability for the design of magnetic bearing controllers.

Multivariate statistical process monitoring methods are often used in chemical process fault diagnosis. In this article, (I) the cycle temporal algorithm (CTA) combined with the dynamic kernel principal component analysis (DKPCA) and the multiway dynamic kernel principal component analysis (MDKPCA) fault detection algorithms are proposed, which are used for continuous and batch process fault detections, respectively. In addition, (II) a fault variable identification model based on reconstructed-based contribution (RBC) model that paves the way for determining the cause of the fault are proposed. The proposed fault diagnosis model was applied to Tennessee Eastman (TE) process and penicillin fermentation process for fault diagnosis. And compare with other fault diagnosis methods. The results show that the proposed method has better detection effects than other methods. Finally, the reconstruction-based contribution (RBC) model method is used to accurately locate the root cause of the fault and determine the fault path.

On-line detection of rotor and load faults in permanent magnet synchronous motors (PMSM) based on spectrum analysis of current or vibration is not capable of identifying the root cause of the fault, since all faults produce identical fault signatures. The sensitivity of fault detection also depends on the motor and controller design making fault detection unreliable. In addition, operation under variable frequency and load limits the effectiveness of spectrum analysis based methods. In this paper, airgap flux monitoring is investigated as an alternative for providing reliable identification of rotor and load faults in PMSMs. Based on the analysis of airgap flux under partial and uniform demagnetization, dynamic eccentricity and load unbalance, an airgap search coil voltage based method for detection and classification of the faults is proposed. The claims made in the paper are verified through experimental testing on an IPMSM under emulated fault conditions along with a comparison to vibration and current spectra-based detection. It is shown that the proposed method provides reliable on-line identification of the faults for cases where conventional spectrum analysis based methods fail.

Intelligent fault detection methods based on deep learning have been developed rapidly in recent years. However, most of these methods are based on supervised learning which requires a fully labeled training set. It is difficult to obtain massive labeled samples in real applications incredibly accurately labeled fault samples from an operating system. The lack of labels and label noise becomes a great challenge for fault detection. To tackle this problem, we propose a Positive-Unlabeled learning based hybrid network (PUHN). It only needs part of the normal operating samples to be labeled. All other samples (including the rest of normal samples and all fault samples) are unlabeled, which greatly reduces the labeling cost. PUHN consists of three modules: a non-negative risk Positive Unlabeled (PU) network for training the classifier, a feature extraction module and a clustering layer for improving data separability and estimating the class priors of PU learning. The three are optimized as a whole and the corresponding optimization strategy is designed. The monitoring data of 24 wind turbines are used to verify the effectiveness and robustness of the proposed method. The experimental results indicate that the proposed method is superior to the benchmark methods, and the performance is significantly better than the supervised learning method when there exists label noise.

Missing data widely exist in industrial processes and lead to difficulties in modelling, monitoring, fault diagnosis, and control. In this paper, we propose a nonlinear method to handle the missing data problem in the offline modelling stage or/and the online monitoring stage of statistical process monitoring. We provide a fast incremental nonlinear matrix completion (FINLMC) method for missing data imputation, which enables us to use kernel methods such as kernel principal component analysis (KPCA) to monitor nonlinear multivariate processes even when there are missing data. We also provide theoretical analysis for the effectiveness of the proposed method. Experiments show that the proposed method can reduce the false alarm rate and improve the fault detection rate in nonlinear processing monitoring with missing data. The proposed method FINLMC can also be used to solve missing data in other problems such as classification and process control.

In this article, we present a method for constructing the fault detector in the form of signal temporal logic (STL) formulas, which can be understood by human users and formally proven to detect faults with probabilistic satisfaction guarantees, for a class of switched nonlinear systems with partially unknown dynamics. First, the partially unknown internal dynamics are approximated by the Gaussian process with stability guarantees. Second, a novel temporal logic inference algorithm is proposed to find the fault detector, which takes advantage of the internal properties of temporal logic and searches for the optimal formula along a partially ordered direction. Moreover, the algorithm is not allowed for missing faults but allowed for false alarms during the temporal logic inference process. In addition, we simulate finitely many trajectories with Chua's circuit and infer the temporal logic formulas with the Gaussian optimization. The results show that the proposed method can find a temporal logic formula to detect the faulty trajectory with a probability guarantee.

The fixed-impedance-based distance relaying schemes find limitation in providing reliable protection to transmission lines connecting doubly fed induction generator based large capacity wind farms and compensated with static synchronous compensator (STATCOM). In this article, the performance of the distance relay protecting such transmission lines is investigated and proposes an advanced protection scheme where the fault detection and fault section identification task is performed using the signs of the superimposed $\alpha$ -aerial components of the measured currents at the line ends and STATCOM location. Furthermore, suitable features are extracted from the relay end measured currents using dual-time transform (TT) and subsequently utilized by decision tree classifier for fault classification. Finally, the single-end traveling wave-based technique is employed using $\alpha$ and $\beta$ -aerial current waves for estimating the fault location. The performance of the proposed protection scheme is evaluated on different fault and nonfault cases generated on a two-bus test power system through MATLAB/Simulink and OPAL-RT (OP4510) manufactured real-time digital simulation environment. The performance of the proposed protection scheme is also compared with some earlier reported works. The demonstrated results justify the effectiveness of the proposed scheme for protecting such complex transmission networks.

Exceptional robust configuration and fault tolerance capabilities make switched reluctance motors (SRMs) a promising technology for automotive applications. Nevertheless, the power transistors in a power converter of an SRM drive are well–considered more vulnerable, compared to the SRM itself. Therefore, accurate and early stage fault detection is crucial for such safety-critical applications. This paper investigates the usage of high-frequency (HF) signal injection for real-time fault identification and localization of power transistors in SRM drives. Unlike the existing diagnostic schemes that often use extra current/voltage sensors and complex algorithms, this simple method injects a sinusoidal HF voltage signal into a standard asymmetrical half-bridge converter-fed SRM. Then the induced HF current is filtered from the fundamental current and its frequency is estimated in real-time for fault localization. Typical power transistor faults (open-circuit and short-circuit) and their location can be effectively detected in the earliest stage by monitoring the variations in the filtered HF signal as well as the fundamental current with the occurrence of the faults. The effectiveness of the proposed diagnostic strategy is validated by thorough simulation and experimental results carried out on a three–phase 12/8 SRM drive system.

This paper addresses the issue of reliability in Industrial Internet of Things (IIoT) in case of missing sensors measurements due to network or hardware problems. We propose to support the fault detection and classification modules, which are the two critical components of a monitoring system for IIoT, with a generative model. The latter is responsible of imputing missing sensor measurements so that the monitoring system performance is robust to missing data. In particular, we adopt Generative Adversarial Networks (GANs) to generate missing sensor measurements and we propose to fine-tune the training of the GAN based on the impact that the generated data have on the fault detection and classification modules. We conduct a thorough evaluation of the proposed approach using the extended Tennessee Eastman Process dataset. Results show that the GAN-imputed data mitigate the impact on the fault detection and classification even in the case of persistently missing measurements from sensors that are critical for the correct functioning of the monitoring system.

Due to the complex muti-branch structure of photovoltaic DC distribution systems, single-ended measurement based protection cannot work well. The DC line-to-line faults are also difficult to locate due to a short time span with fault characteristics that are affected by various converter structures and control algorithms. To resolve these issues, a new DC line protection based on active converter injection is proposed. This method uses communication cooperation to realize the coordination of protection and converter control, then the converter becomes a characteristic signal injection source, that can build a clear protection boundary. The proposed method can complete fast fault detection and precise fault location. Moreover, compared with current DC protection techniques, it does not require additional injection equipment and high sampling frequency. Hardware tests are carried out, illustrating that the proposed approach is effective to achieve line protection through active control of power electronics.

A receiver autonomous integrity monitoring (RAIM) framework for ground vehicle navigation using ambient cellular signals of opportunity (SOPs) and an inertial measurement unit (IMU) is developed. The proposed framework accounts for two types of errors that compromise the integrity of the navigation solution: (i) multipath and (ii) unmodeled biases in the cellular pseudorange measurements due to line-of-sight (LOS) signal blockage and high signal attenuation. This paper, first, characterizes the multipath in a cellular-based navigation framework. Next, a fault detection and exclusion technique for a cellular-based navigation framework is developed. Simulation and experimental results with real long-term evolution (LTE) signals are presented evaluating the efficacy of the proposed RAIM-based fault detection and exclusion technique on a ground vehicle navigating in a deep urban environment in the absence of global navigation satellite system (GNSS) signals. The experimental results on a ground vehicle traversing 825 m in an urban environment show that the proposed RAIM-based measurement exclusion technique reduces the position root mean-squared error (RMSE) by 66%.

Satisfying fault ride-through requirements, solar photovoltaic (PV) plant remains connected to the grid even during a fault. With current limiting feature, the plant current during a fault becomes comparable to its rated current. This imposes challenges to conventional local data-based protection methods employed in the relay on PV-side. In such a situation, the selectivity of a protection scheme is crucial for maintaining an uninterrupted supply to customers. In this article, a q-axis component-based protection scheme is proposed for the distribution system with a PV plant. The method extracts the second-order harmonic components of q-axis voltage and current at relaying point for fault detection and direction identification. Using this direction information at both ends of a line segment to be protected, internal and external faults are discriminated. The performance of the proposed method is tested on 14-bus distribution system for various fault cases. A comparative analysis with existing techniques shows the strength of the proposed method. Hardware-in-loop testing using OPAL-RT simulator as system and Arduino NANO microcontroller as relay validates the performance of the proposed method in real-time.