Showing papers on "Fault detection and isolation published in 2017"

A Convolutional Neural Network for Fault Classification and Diagnosis in Semiconductor Manufacturing Processes

Abstract: Many studies on the prediction of manufacturing results using sensor signals have been conducted in the field of fault detection and classification (FDC) for semiconductor manufacturing processes. However, fault diagnosis used to find clues as to root causes remains a challenging area. In particular, process monitoring using neural networks has been employed to only a limited extent because it is a black box model, making the relationships between input data and output results difficult to interpret in actual manufacturing settings, despite its high classification performance. In this paper, we propose a convolutional neural network (CNN) model, named FDC-CNN, in which a receptive field tailored to multivariate sensor signals slides along the time axis, to extract fault features. This approach enables the association of the output of the first convolutional layer with the structural meaning of the raw data, making it possible to locate the variable and time information that represents process faults. In an experiment on a chemical vapor deposition process, the proposed method outperformed other deep learning models.

Event-Triggered Fault Detection of Nonlinear Networked Systems

Abstract: This paper investigates the problem of fault detection for nonlinear discrete-time networked systems under an event-triggered scheme. A polynomial fuzzy fault detection filter is designed to generate a residual signal and detect faults in the system. A novel polynomial event-triggered scheme is proposed to determine the transmission of the signal. A fault detection filter is designed to guarantee that the residual system is asymptotically stable and satisfies the desired performance. Polynomial approximated membership functions obtained by Taylor series are employed for filtering analysis. Furthermore, sufficient conditions are represented in terms of sum of squares (SOSs) and can be solved by SOS tools in MATLAB environment. A numerical example is provided to demonstrate the effectiveness of the proposed results.

A Data-Driven Fault Diagnosis Methodology in Three-Phase Inverters for PMSM Drive Systems

Abstract: Permanent magnet synchronous motor and power electronics-based three-phase inverter are the major components in the modern industrial electric drive system, such as electrical actuators in an all-electric subsea Christmas tree. Inverters are the weakest components in the drive system, and power switches are the most vulnerable components in inverters. Fault detection and diagnosis of inverters are extremely necessary for improving drive system reliability. Motivated by solving the uncertainty problem in fault diagnosis of inverters, which is caused by various reasons, such as bias and noise of sensors, this paper proposes a Bayesian network-based data-driven fault diagnosis methodology of three-phase inverters. Two output line-to-line voltages for different fault modes are measured, the signal features are extracted using fast Fourier transform, the dimensions of samples are reduced using principal component analysis, and the faults are detected and diagnosed using Bayesian networks. Simulated and experimental data are used to train the fault diagnosis model, as well as validate the proposed fault diagnosis methodology.

Deep Learning Enabled Fault Diagnosis Using Time-Frequency Image Analysis of Rolling Element Bearings

Abstract: Traditional feature extraction and selection is a labor-intensive process requiring expert knowledge of the relevant features pertinent to the system. This knowledge is sometimes a luxury and could introduce added uncertainty and bias to the results. To address this problem a deep learning enabled featureless methodology is proposed to automatically learn the features of the data. Time-frequency representations of the raw data are used to generate image representations of the raw signal, which are then fed into a deep convolutional neural network (CNN) architecture for classification and fault diagnosis. This methodology was applied to two public data sets of rolling element bearing vibration signals. Three time-frequency analysis methods (short-time Fourier transform, wavelet transform, and Hilbert-Huang transform) were explored for their representation effectiveness. The proposed CNN architecture achieves better results with less learnable parameters than similar architectures used for fault detection, including cases with experimental noise.

DC Fault Detection and Location in Meshed Multiterminal HVDC Systems Based on DC Reactor Voltage Change Rate

Abstract: The change rate of the dc reactor voltage with predefined protection voltage thresholds is proposed to provide fast and accurate dc fault detection in a meshed multiterminal HVDC system. This is equivalent to the measurement of the second derivative of the dc current but has better robustness in terms of electromagnetic-interference noise immunization. In addition to fast dc fault detection, the proposed scheme can also accurately discriminate the faulty branch from the healthy ones in a meshed dc network by considering the voltage polarities and amplitudes of the two dc reactors connected to the same converter dc terminal. Fast fault detection leads to lower fault current stresses on dc circuit breakers and converter equipment. The proposed method requires no telecommunication, is independent of power-flow direction, and is robust to fault resistance variation. Simulation of a meshed three-terminal HVDC system demonstrates the effectiveness of the proposed dc fault detection scheme.

Active Steering Actuator Fault Detection for an Automatically-Steered Electric Ground Vehicle

Abstract: In this paper, we investigate the actuator fault detector design problem for an electric ground vehicle (EGV) that is equipped with an active front-wheel steering system. Since the EGV can be steered by a motor automatically, it is desired to design a fault detector for the steering actuator for safety reasons. A two degree of freedom lateral nonlinear vehicle model is established. The nonlinear vehicle model is converted to a linear-parameter-varying (LPV) form and the scheduling vector is related to the vehicle longitudinal velocity. Since it is not easy to measure the longitudinal velocity precisely, the uncertain measurement on the longitudinal velocity is considered and the weighting factors of LPV submodels are subject to uncertainties. Based on the uncertain LPV model, a gain-scheduling fault detector is proposed and an augmented system is obtained. The desired steering angle and the faulty steering angle are both involved in the augmented system. As the steering angle generally has a low-frequency working range, the steering angle amplitude spectrums of three different maneuvers are studied, and the frequency working range is determined. The stability, the $\mathcal {H}{\_}$ performance, and the $\mathcal {H}_{\infty }$ performance of the augmented system are all exploited. Based on the analysis results, the mixed $\mathcal {H}{\_}$ / $\mathcal {H}_{\infty }$ fault detector design method is developed. An experimental test is used to show the performance of the designed fault detector.

Distributed Parallel PCA for Modeling and Monitoring of Large-Scale Plant-Wide Processes With Big Data

Abstract: In order to deal with the modeling and monitoring issue of large-scale industrial processes with big data, a distributed and parallel designed principal component analysis approach is proposed. To handle the high-dimensional process variables, the large-scale process is first decomposed into distributed blocks with a priori process knowledge. Afterward, in order to solve the modeling issue with large-scale data chunks in each block, a distributed and parallel data processing strategy is proposed based on the framework of MapReduce and then principal components are further extracted for each distributed block. With all these steps, statistical modeling of large-scale processes with big data can be established. Finally, a systematic fault detection and isolation scheme is designed so that the whole large-scale process can be hierarchically monitored from the plant-wide level, unit block level, and variable level. The effectiveness of the proposed method is evaluated through the Tennessee Eastman benchmark process.

Automated fault detection without seismic processing

Abstract: For hydrocarbon exploration, large volumes of data are acquired and used in physical modeling-based workflows to identify geologic features of interest such as fault networks, salt bodies, or, in general, elements of petroleum systems. The adjoint modeling step, which transforms the data into the model space, and subsequent interpretation can be very expensive, both in terms of computing resources and domain-expert time. We propose and implement a unique approach that bypasses these demanding steps, directly assisting interpretation. We do this by training a deep neural network to learn a mapping relationship between the data space and the final output (particularly, spatial points indicating fault presence). The key to obtaining accurate predictions is the use of the Wasserstein loss function, which properly handles the structured output — in our case, by exploiting fault surface continuity. The promising results shown here for synthetic data demonstrate a new way of using seismic data and suggest more d...

Transient-Voltage-Based Protection Scheme for DC Line Faults in the Multiterminal VSC-HVDC System

Abstract: DC line faults are major issues for a multiterminal high-voltage direct current (HVDC) system based on voltage-source converter (VSC). The fault current increases quickly along with a large peak, and complete isolation of the faulted system is not a viable option. Therefore, protection with high selectivity and accuracy is essential. In this paper, a new protection scheme for dc line in multiterminal VSC-HVDC system is proposed, which consists of a main protection and a backup protection. Both the protection principles are based on the supplemental inductor placed at each end of the dc line. Fault identification can be achieved by calculating the ratio of the transient voltages (ROTV) at both sides of the inductor. The main protection is able to detect the fault quickly without communication, while the backup protection is a pilot method based on the ROTVs at both ends of dc line, which is employed to identify the high-resistance faults and offer a backup in case the former fails. Comparison with some previous protection methods shows that the performance of the proposed protection scheme is promising. Numerous simulation studies carried out in PSCAD/EMTDC and real-time digital simulator (RTDS) under various conditions have demonstrated that fault identification with high selectivity and strong robustness against fault resistance and disturbance can be achieved by employing the proposed protection scheme.

Dependable Structural Health Monitoring Using Wireless Sensor Networks

Abstract: As an alternative to current wired-based networks, wireless sensor networks (WSNs) are becoming an increasingly compelling platform for engineering structural health monitoring (SHM) due to relatively low-cost, easy installation, and so forth. However, there is still an unaddressed challenge: the application-specific dependability in terms of sensor fault detection and tolerance. The dependability is also affected by a reduction on the quality of monitoring when mitigating WSN constrains (e.g., limited energy, narrow bandwidth). We address these by designing a dependable distributed WSN framework for SHM (called DependSHM ) and then examining its ability to cope with sensor faults and constraints. We find evidence that faulty sensors can corrupt results of a health event (e.g., damage) in a structural system without being detected. More specifically, we bring attention to an undiscovered yet interesting fact, i.e., the real measured signals introduced by one or more faulty sensors may cause an undamaged location to be identified as damaged (false positive) or a damaged location as undamaged (false negative) diagnosis. This can be caused by faults in sensor bonding, precision degradation, amplification gain, bias, drift, noise, and so forth. In DependSHM , we present a distributed automated algorithm to detect such types of faults, and we offer an online signal reconstruction algorithm to recover from the wrong diagnosis. Through comprehensive simulations and a WSN prototype system implementation, we evaluate the effectiveness of DependSHM .

Fault Detection, Isolation, and Service Restoration in Distribution Systems: State-of-the-Art and Future Trends

Abstract: This paper surveys the conceptual aspects, as well as recent developments in fault detection, isolation, and service restoration (FDIR) following an outage in an electric distribution system. This paper starts with a discussion of the rationale for FDIR, and then investigates different areas of the FDIR problem. Recently reported approaches are compared and related to discussions on current practices. This paper then addresses some of the often-cited associated technical, environmental, and economic challenges of implementing self-healing for the distribution grid. The review concludes by pointing toward the need and directions for future research.

Sensor Fault Classification Based on Support Vector Machine and Statistical Time-Domain Features

Abstract: This paper deals with the problem of fault detection and diagnosis in sensors considering erratic, drift, hard-over, spike, and stuck faults. The data set containing samples of the abovementioned fault signals was acquired as follows: normal data signals were obtained from a temperature-to-voltage converter by using an Arduino Uno microcontroller board and MATLAB. Then, faults were simulated in normal data to get 100 samples of each fault, in which one sample is composed of 1000 data elements. A support vector machine (SVM) was used for data classification in a one-versus-rest manner. The statistical time-domain features, extracted from a sample, were used as a single observation for training and testing SVM. The number of features varied from 5 to 10 to examine the effect on accuracy of SVM. Three different kernel functions used to train SVM include linear, polynomial, and radial-basis function kernels. The fault occurrence event in fault samples was chosen randomly in some cases to replicate a practical scenario in industrial systems. The results show that an increase in the number of features from 5 to 10 hardly increases the total accuracy of the classifier. However, using ten features gives the highest accuracy for fault classification in an SVM. An increase in the number of training samples from 40 to 60 caused an overfitting problem. The $k$ -fold cross-validation technique was adopted to overcome this issue. The increase in number of data elements per sample to 2500 increases the efficiency of the classifier. However, an increase in the number of training samples to 400 reduces the capability of SVM to classify stuck fault. The receiver operating characteristics curve comparison shows the efficiency of SVM over a neural network.

Fault Detection for Photovoltaic Systems Based on Multi-Resolution Signal Decomposition and Fuzzy Inference Systems

Abstract: This paper presents a detection scheme for DC side short-circuit faults of photovoltaic (PV) arrays that consist of multiple PV panels connected in a series/parallel configuration. Such faults are nearly undetectable under low irradiance conditions, particularly, when a maximum power point tracking algorithm is in-service. If remain undetected, these faults can considerably lower the output energy of solar systems, damage the panels, and potentially cause fire hazards. The proposed fault detection scheme is based on a pattern recognition approach that employs a multiresolution signal decomposition technique to extract the necessary features, based on which a fuzzy inference system determines if a fault has occurred. The presented case studies (both simulation and experimental) demonstrate the effective and reliable performance of the proposed method in detecting PV array faults.

Line-to-Line Fault Detection for Photovoltaic Arrays Based on Multiresolution Signal Decomposition and Two-Stage Support Vector Machine

Abstract: Fault detection in photovoltaic (PV) arrays becomes difficult as the number of PV panels increases Particularly, under low irradiance conditions with an active maximum power point tracking algorithm, line-to-line (L-L) faults may remain undetected because of low fault currents, resulting in loss of energy and potential fire hazards This paper proposes a fault detection algorithm based on multiresolution signal decomposition for feature extraction, and two-stage support vector machine (SVM) classifiers for decision making This detection method only requires data of the total voltage and current from a PV array and a limited amount of labeled data for training the SVM Both simulation and experimental case studies verify the accuracy of the proposed method

A Robust Observer-Based Method for IGBTs and Current Sensors Fault Diagnosis in Voltage-Source Inverters of PMSM Drives

Abstract: Permanent magnet synchronous motors (PMSMs) drives using three-phase voltage-source inverters (VSIs) are currently used in many industrial applications. The reliability of VSIs is one of the most important factors to improve the reliability and availability levels of the drive. Accordingly, this paper presents a robust fault diagnostic method for multiple insulated gate bipolar transistors (IGBTs) open-circuit faults and current sensor faults in three-phase PMSM drives. The proposed observer-based algorithm relies on an adaptive threshold for fault diagnosis. Current sensor and open-circuit faults can be distinguished and the faulty sensors and/or power semiconductors are effectively isolated. The proposed technique is robust to machine parameters and load variations. Several simulation and experimental results using a vector-controlled PMSM drive are presented, showing the diagnostic algorithm robustness against false alarms and its effectiveness in both IGBTs and current sensors fault diagnosis.

Neural adaptive observer-based sensor and actuator fault detection in nonlinear systems: Application in UAV.

Abstract: A new online detection strategy is developed to detect faults in sensors and actuators of unmanned aerial vehicle (UAV) systems. In this design, the weighting parameters of the Neural Network (NN) are updated by using the Extended Kalman Filter (EKF). Online adaptation of these weighting parameters helps to detect abrupt, intermittent, and incipient faults accurately. We apply the proposed fault detection system to a nonlinear dynamic model of the WVU YF-22 unmanned aircraft for its evaluation. The simulation results show that the new method has better performance in comparison with conventional recurrent neural network-based fault detection strategies.

Model-Based Fault Detection and Identification for Switching Power Converters

Abstract: We present the analysis, design, and experimental validation of a model-based fault detection and identification (FDI) method for switching power converters using a model-based state estimator approach. The proposed FDI approach is general in that it can be used to detect and identify arbitrary faults in components and sensors in a broad class of switching power converters. The FDI approach is experimentally demonstrated on a nanogrid prototype with a 380-V dc distribution bus. The nanogrid consists of four different switching power converters, including a buck converter, an interleaved boost converter, a single-phase rectifier, and a three-phase inverter. We construct a library of fault signatures for possible component and sensor faults in all four converters. The FDI algorithm successfully achieves fault detection in under 400 $\mu$ s and fault identification in under 10 ms for faults in each converter. The proposed FDI approach enables a flexible and scalable solution for improving fault tolerance and awareness in power electronics systems.

Bearing Fault Detection by One-Dimensional Convolutional Neural Networks

Abstract: Bearing faults are the biggest single source of motor failures. Artificial Neural Networks (ANNs) and other decision support systems are widely used for early detection of bearing faults. The typical decision support systems require feature extraction and classification as two distinct phases. Extracting fixed features each time may require a significant computational cost preventing their use in real-time applications. Furthermore, the selected features for the classification phase may not represent the most optimal choice. In this paper, the use of 1D Convolutional Neural Networks (CNNs) is proposed for a fast and accurate bearing fault detection system. The feature extraction and classification phases of the bearing fault detection are combined into a single learning body with the implementation of 1D CNN. The raw vibration data (signal) is fed into the proposed system as input eliminating the need for running a separate feature extraction algorithm each time vibration data is analyzed for classification. Implementation of 1D CNNs results in more efficient systems in terms of computational complexity. The classification performance of the proposed system with real bearing data demonstrates that the reduced computational complexity is achieved without a compromise in fault detection accuracy.

Adaptively Adjusted Event-Triggering Mechanism on Fault Detection for Networked Control Systems

Abstract: This paper studies the problem of adaptively adjusted event-triggering mechanism-based fault detection for a class of discrete-time networked control system (NCS) with applications to aircraft dynamics. By taking into account the fault occurrence detection progress and the fault occurrence probability, and introducing an adaptively adjusted event-triggering parameter, a novel event-triggering mechanism is proposed to achieve the efficient utilization of the communication network bandwidth. Both the sensor-to-control station and the control station-to-actuator network-induced delays are taken into account. The event-triggered sensor and the event-triggered control station are utilized simultaneously to establish new network-based closed-loop models for the NCS subject to faults. Based on the established models, the event-triggered simultaneous design of fault detection filter (FDF) and controller is presented. A new algorithm for handling the adaptively adjusted event-triggering parameter is proposed. Performance analysis verifies the effectiveness of the adaptively adjusted event-triggering mechanism, and the simultaneous design of FDF and controller.

Centralized protection strategy for medium voltage DC microgrids

Abstract: This paper presents a centralized protection strategy for medium voltage dc microgrids. The proposed strategy consists of a communication-assisted fault detection method with a centralized protection coordinator and a fault isolation technique that provides an economic, fast, and selective protection by using the minimum number of dc circuit breakers. The proposed method is also supported by a backup protection that is activated if communication fails. This paper also introduces a centralized self-healing strategy that guarantees successful operation of zones that are separated from the main grid after the operation of the protection devices. Furthermore, to provide a more reliable protection, thresholds of the protection devices are adapted according to the operational modes of the microgrid and the status of distributed generators. The effectiveness of the proposed protection strategy is validated through real-time simulation studies based on the hardware in the loop approach.