Fault Detection Phase Research Articles

Fault-Tolerant Training Enabled by On-Line Fault Detection for RRAM-Based Neural Computing Systems

An resistive random-access memory (RRAM)-based computing system (RCS) is an attractive hardware platform for implementing neural computing algorithms. On-line training for RCS enables hardware-based learning for a given application and reduces the additional error caused by device parameter variations. However, a high occurrence rate of hard faults due to immature fabrication processes and limited write endurance restrict the applicability of on-line training for RCS. We propose a fault-tolerant on-line training method that alternates between a fault-detection phase and a fault-tolerant training phase. In the fault-detection phase, a quiescent-voltage comparison method is utilized. In the training phase, a threshold-training method and a remapping scheme is proposed. Our results show that, compared to neural computing without fault tolerance, the recognition accuracy for the Cifar-10 dataset improves from 37% to 83% when using low-endurance RRAM cells, and from 63% to 76% when using RRAM cells with high endurance but a high percentage of initial faults.

A complete diagnosis of faulty sensor modules in a wireless sensor network

Wireless sensor network (WSN) is a collection of sensor devices, which collects information from the environment and sends to the base station by single-hop or multi-hop communication towards achieving some predefined objectives. Due to the multi-functional applications in the adventurer environment, the sensor modules are damaged by various external and internal sources, which leads to a failure sensor network. Therefore, automated fault diagnosis and fault tolerance are of utmost importance in WSN. In this article, a complete fault diagnosis methodology is proposed for faulty sensors in the failed WSN. The proposed methodology consists of four phases of functioning such as initialization, fault detection, fault classification, and fault tolerance phase. The hard fault and link fault are detected by checksum and Fletcher’s checksum method. The soft faults (permanent, intermittent, and transient) are detected by Mann-Whitney U statistical test. The Gaussian transformation function with a classification factor is used for soft faults classification. The fault tolerance is attained in the sensor system by the stepwise regressional method. The proposed methodology is implemented in a real testbed experiment with simulation setup. The parameters like fault detection accuracy, false alarm rate, false positive rate, fault classification rate, false classification rate, R-squared value, standard error, fault probability, and degree of sensors are considered for the evaluation of the proposed diagnosis methodology.

Introduction to the Special Section on Intelligent Fault Monitoring and Fault–Tolerant Control in Power Electronics, Drives and Renewable Energy Systems

Abstract This article constitutes an introductory part of the special section on Intelligent Fault Monitoring and Fault-Tolerant Control in Power Electronics, Drives and Renewable Energy Systems. In the current issue of the journal, the first part of this section is published. Accepted articles are focussed mainly on the sensor-fault diagnosis methods for T-type inverter-fed dual-three phase PMSM drives, partial demagnetization, faults of the permanent magnet synchronous generator (PMSG) and online open phase fault detection (FD) in the sensorless five-phase induction motor drive implemented with an inverter output LC filter and third harmonic injection. Also, neural networks (NN) application in the detection of stator and rotor electrical faults of induction motors has been proposed in one of the papers, and the observer-based FD concept for unknown systems using input–output measurements was applied to a brushless direct current motor drive with unknown parameters.

Impedance differential plane for fault detection and faulty phase identification of FACTS compensated transmission line

Impedance differential plane for fault detection and faulty phase identification of FACTS compensated transmission line

Order-Information-Based Phase Partition and Fault Detection for Batch Processes

Batch processes have significantly different properties across different phases. It is essential to divide batch processes reasonably and model corresponding phases correctly for fault detection. Order information is critical for phase partition. Different from traditional phase partition methods, which customarily adopt the results of PCA for advanced research, a novel method called Markov-chain-based spectral partition (MCSP) is proposed in this paper, which takes the sequence (order) information into consideration for phase partition. Due to the combination of neighborhood information and order information, MCSP can separate each batch into major phases automatically. The time-varying characteristics remain relatively stable in each phase, which is suitable for following a Gaussian-mixture-model-based fault detection model to reflect the phase properties. Due to its simple and intuitive format, our proposed method has superior performance in fault detection. The effectiveness of MCSP is illustrated in ...

Online Open-Phase Fault Detection for Permanent Magnet Machines With Low Fault Harmonic Magnitudes

Conventional permanent magnet (PM) machine open-phase fault detection relies on the estimation of current harmonics using available phase current sensors. Because the magnitudes of current harmonic are proportional to the machine load condition, it is still a challenge to detect a phase fault at light load when fault-induced current harmonics are too small to estimate. This paper investigates the phase fault detection specifically for the conditions under low fault harmonic magnitudes. Both fault-induced harmonics in dq currents and neutral point (NP) voltage are analyzed to find a suited signal for the fault detection. By adding a voltage sensor to measure the NP voltage in a PM machine, a better fault detection performance is achieved because the fault harmonic in NP voltage results in the reduced effect on torque loads and sensor measurement errors. A 50-W PM machine is tested to verify the phase fault detection performance under low fault harmonic magnitudes. This paper includes the experimental evaluation on the detection limitation in terms of load capability using current and voltage fault signals.

Multifault diagnosis in WSN using a hybrid metaheuristic trained neural network

Wireless sensor networks are susceptible to failures of nodes and links due to various physical or computational reasons. Some physical reasons include a very high temperature, a heavy load over a node, and heavy rain. Computational reasons could be a third-party intrusive attack, communication conflicts, or congestion. Automated fault diagnosis has been a well-studied problem in the research community. In this paper, we present an automated fault diagnosis model that can diagnose multiple types of faults in the category of hard faults and soft faults. Our proposed model implements a feed-forward neural network trained with a hybrid metaheuristic algorithm that combines the principles of exploration and exploitation of the search space. The proposed methodology consists of different phases, such as a clustering phase, a fault detection and classification phase, and a decision and diagnosis phase. The implemented methodology can diagnose composite faults, such as hard permanent, soft permanent, intermittent, and transient faults for sensor nodes as well as for links. The proposed implementation can also classify different types of faulty behavior for both sensor nodes and links in the network. We present the obtained theoretical results and computational complexity of the implemented model for this particular study on automated fault diagnosis. The performance of the model is evaluated using simulations and experiments conducted using indoor and outdoor testbeds.

Monitoring of Wastewater Treatment Plants using Improved Univariate Statistical Strategy

In this paper, we propose an enhanced monitoring of wastewater treatment plant (WWTP) using state estimation-based fault detection strategies. The WWTP state estimation problem is performed using particle filter (PF) technique. The PF has shown good improvement and provides a significant advantage over extended Kalman filter (EKF), unscented Kalman filter (UKF) techniques and can be applied to nonlinear models with non-Gaussian errors. The fault detection phase is achieved using a novel chart called multiscale exponentially weighted moving average chart (MS-EWMA). The new chart allows to optimize the smoothing parameter (λ) and control width L of EWMA chart to deal with the dynamic nature of WWTPs. It enables also to extract accurate deterministic features and decorrelate autocorrelated measurements using dynamical multiscale representation. The fault detection performance is studied using simulated COST wastewater treatment BSM1 model. The BSM1, provided by the IWA Task Group on Benchmarking of Control Strategies, is a simulation platform that allows for creating sensor faults disturbances in a wastewater treatment plant. The developed technique is used to enhance fault detection of the BSM1 system through monitoring some of the key variables involved in this model. The results demonstrate the effectiveness of the proposed PF-based MS-optimized EWMA method over EWMA and Shewhart charts.

Time Sequential Phase Partition and Modeling Method for Fault Detection of Batch Processes

Different operation phases in batch processes cover distinguishing behaviors, so establishing statistical models for each identical phase become an effective way for batch monitoring. In this paper, a new adaptive phase partition and online fault detection method is proposed, which can track the phase’s transition by time sequence and has less reliance on parameters’ selection. The discussion and analysis of this proposed method follows. In this proposed method, the information contained in every sample time will be evaluated, and the change tendency of feature is demonstrated on a batch prospect. Then, two control bounds are designed for the feature tendency, the stable, and the transitional phases that have a different feature level and play certainly roles in process operation, will be identified automatically. For online monitoring, the new fault detection strategy is composed of modeling the PCA and PLS statistical methods for each identified phase, three statistics are established to ensure the data-decomposing reliable. The proposed method is applied to the industrial penicillin fermentation process, and the experimental result shows better performance in phase partition and fault detection.

Heterogeneous fault diagnosis for wireless sensor networks

Fault diagnosis has been considered as a very challenging problem in wireless sensor network (WSN) research. Faulty nodes having different behavior such as hard, soft, intermittent, and transient fault are called as heterogeneous faults in wireless sensor networks. This paper presents a heterogeneous fault diagnosis protocol for wireless sensor networks. The proposed protocol consists of three phases, such as clustering phase, fault detection phase, and fault classification phase to diagnose the heterogeneous faulty nodes in the wireless sensor networks. The protocol strategy is based on time out mechanism to detect the hard faulty nodes, and analysis of variance method (ANOVA test) to detect the soft, intermittent, and transient faulty nodes in the network. The feed forward probabilistic neural network (PNN) technique is used to classify the different types of faulty nodes in the network. The performance of the proposed heterogeneous fault diagnosis protocol is evaluated using network simulator NS-2.35. The evaluation of the proposed model is also carried out by the testbed experiment in an indoor laboratory environment and outdoor environment.

An effective graph‐theoretic approach towards simultaneous detection of fault(s) and cut(s) in wireless sensor networks

SummaryA failed sensor node partitions a wireless sensor network (WSN) into 2 or more disjoint components, which is called as a cut in the network. The cut detection has been considered as a very challenging problem in the WSN research. In this paper, we propose a graph‐theoretic distributed protocol to detect simultaneously the faults and cuts in the WSN. The proposed approach could be accomplished mainly in 3 phases, such that initialization phase, fault detection phase, and a cut detection phase. The protocol is an iterative method where at every time iteration, the node updates its state to calculate the potential factor. We introduced 2 terminologies such as a safe zone or cut zone of the network. The proposed method has been evaluated regarding various performance evaluation measures by implementing the same in the network simulator NS–2.35. The obtained results show that the proposed graph‐theoretic approach is simple yet very powerful for the intended tasks.

On-Line Stator Open-Phase Fault Detection and Tolerant Control for Permanent Magnet Machines Using the Neutral Point Voltage

Conventional machine stator open-phase fault detection methods rely on the detection of current harmonics from the sensing of phase currents. Because the magnitudes of current harmonics are proportional to the machine load condition, it is a challenge for the phase fault detection at the light load condition when fault-induced current harmonics are too small to detect. This paper proposes an improved stator phase fault detection for permanent magnet (PM) machines based on the use of neutral point (NP) in Y-connected windings. It is shown that the first-order voltage harmonic is resultant once the open-phase fault occurs. Comparing with prior detection methods based on the current measurement, the fault detection using NP voltage is insensitive to the load condition, because the proposed fault signal is induced by the voltage unbalance as a result of the open phase. In addition, considering the fault tolerant control, a single-phase drive is developed by connecting the machine NP to one of the inverter legs. The machine can drive at the most efficient condition using standard three-leg inverters under phase fault. A 50-W PM machine with the accessible NP is used to evaluate the proposed open-phase fault detection and tolerant control method.

Composite Fault Diagnosis in Wireless Sensor Networks Using Neural Networks

Wireless sensor networks (WSNs) are spatially distributed devices to support various applications. The undesirable behavior of the sensor node affects the computational efficiency and quality of service. Fault detection, identification, and isolation in WSNs will increase assurance of quality, reliability, and safety. In this paper, a novel neural network based fault diagnosis algorithm is proposed for WSNs to handle the composite fault environment. Composite fault includes hard, soft, intermittent, and transient faults. The proposed fault diagnosis protocol is based on gradient descent and evolutionary approach. It detects, diagnose, and isolate the faulty nodes in the network. The proposed protocol works in four phases such as clustering phase, communication phase, fault detection and classification phase, and isolation phase. Simulation results show that the proposed protocol performs better than the existing protocols in terms of detection accuracy, false alarm rate, false positive rate, and detection latency.

Intelligent diagnosis of bearing knock faults in internal combustion engines using vibration simulation

Big-end bearing knock faults in IC engines can be considered as a real industrial case of a slider-crank mechanism including a joint with clearance and lubrication. In this paper, an Artificial Neural Network (ANN) based system was used to solve the problem of intelligent big-end bearing knock fault diagnosis in Internal Combustion (IC) engine. But when the ANN is used in machine condition monitoring, it is either unlikely or uneconomical to experience all different real faults to generate sufficient training data. Therefore, model based method should be a viable way to generate adequate data to train the networks for the intelligent big-end bearing fault diagnosis in IC engines. In order to evaluate and update the simulation model, experiments with normal bearing clearance and with different oversize bearing clearances were first carried out on the engine test rig. It was found that the relevant diagnostic information lies in the squared envelope of the vibration signals. Therefore, we only need build a proper simulation model to simulate the correct envelope signals rather than the raw vibration signals. As the important inputs of the simulation model, the inertia properties of the simulated engine components were also measured and studied. Next, we built an ANN-based bearing knock diagnosis system which consists of three phases: fault detection phase, fault localization phase and fault severity identification phase. Particularly, a saturating linear function is selected as the transfer function of the fault severity identification stage, so the networks can linearly classify the fault levels and the output is more in agreement with the reality in industry. Following the feature extraction and selection from the processed squared envelope signals, the networks were purely trained by the simulated data with normal bearing clearance and with different oversize bearing clearances. Finally the networks was tested by the real experimental data and it was demonstrated that the networks can successfully detect different bearing knock faults in real tests, and also classify the faults' location and severity levels.

Redundancy and lack of information in specific problems of GNSS measurement processing

The paper focuses on four problems of GNSS measurement processing. The difficulty of problems lies in large volume of available data and limited resources of onboard equipment (phase measurement ambiguity resolution, fault detection and isolation); or in insufficient volume and fragmentarity of data (multiantenna heading determination in constrained conditions, navigation of geostationary satellites). Theoretical foundations of proposed solutions are given. Experimental results are provided.

Fault detection and isolation of bearings in a drive reducer of a hot steel rolling mill

Defective bearings are a major concern in rotating machinery. In this work we propose a two-step scheme, relying on two complementary data-driven techniques, for fault detection and isolation for a drive reducer in a hot steel rolling mill. A preliminary fault detection phase is based on a computationally lightweight time-domain multivariate statistical technique. Secondly, a more computationally intensive frequency-domain analysis method is used to confirm the fault detection and provide information on its frequency characteristics. Automatic procedures are sketched for the application of both techniques. Bearing defect models are employed to test their fault detection and isolation capabilities.

D-Q Current Signature-Based Faulted Phase Localization for SM-PMAC Machine Drives

Fault diagnostic algorithms in motor drives prevent damage to the motor, drive, and overall system by providing an additional level of safety. A single phase open (SPO) fault in a field-oriented controlled drive is one of the perilous faults that require immediate detection due to resulting oscillatory electromagnetic torque. SPO fault results in a significantly high electromagnetic torque ripple. The rapid mechanical vibration caused by the torque ripple exposes the system to failure modes. The single open phase fault detection algorithm proposed in this paper enables the rapid detection of the fault, with minimal system resources, compared to existing methods. Proof of accurate fault detection capability for a wide speed range is presented through simulation and experimental results. The proposed method is also capable of localizing the faulted phase via a novel time domain rotor reference frame current signature analysis.

Shunt Petersen-Coil with Resistor for Single Phase Fault Detection in Resonant Grounded Power Distribution Systems

The failure current in resonant grounder power distribution system is small, so it is difficult to detect the fault feeder. This passage presents the equivalent circuit of resonant grounded system, and discusses the difference of electrical characteristics between faulty feeder and sound feeders by using shunt resistors. To reduce the influence of shunt resistors on the system and improve the detection sensitivity, it presents the method of shunting multi-level resistors, and it proves the sensitivity and reliability of this method by EMTP simulation.

Fault diagnosis and fault‐tolerant control for non‐Gaussian non‐linear stochastic systems using a rational square‐root approximation model

The purpose of the fault detection and diagnosis of stochastic distribution control systems is to use the measured input and the system output probability density functions (PDFs) to obtain the fault information of the system. In this paper, the rational square-root B-spline model is used to represent the dynamics between the output PDF and the input. This is then followed by the novel design of a non-linear neural network observer-based fault diagnosis (FD) algorithm so as to diagnose the fault in the dynamic part of such systems. Convergency analysis is performed for the error dynamic system raised from the fault detection and diagnosis phase using the Lyapunov stability theorem. Finally, based on the FD information, a new fault-tolerant control based on proportional integral tracking control scheme is designed to make the post-fault PDF still track the given distribution. A simulated example is given to illustrate the efficiency of the proposed algorithms.

사용자 다이나믹을 고려한 GPS 반송파 고장검출

본 논문은 동적환경의 GPS RTK (Global Positioning System Real Time Kinematic) 수행을 위한 반송파 고장검출에 대해 기술한다. 다양한 고장요소에 노출된 동적수신환경에서 신뢰성 높은 반송파를 확보하기 위해 신호변화 요인을 분리하여 이상 유/무를 판별하는 기법을 사용하였다. 신호변화 요소는 사용자 다이나믹, 위성 다이나믹, 오차요인으로 분리하였으며, 상대적으로 불규칙한 사용자 다이나믹 추정 성능 향상을 위해 미지정수가 소거된 시간차분 반송파와 도플러를 혼합하여 사용하였다. 시뮬레이션을 통해 제안한 검출 기법의 성능을 검증하였으며, 임펄스, 스텝, 램프 고장에 대한 검출 가능성을 분석하였다. This paper presents a Carrier phase fault detection (FD) method for GPS RTK (Global Positioning System Real Time Kinematic) in dynamic environment. There are various error sources in dynamic environment and these errors decrease the reliability of FD results. Due to the reason, Carrier phase measurements are separated into satellite induced signal, user induced signal and other remaining errors. Especially the user-induced signal is computed by user dynamic which is estimated by time-differenced Carrier phase (TDCP) and Doppler shift. TDCP makes it possible to avoid integer ambiguity resolution. Computer simulation is conducted to verify the suggested method. By applying impulse, step and ramp faults, the FD performance is analyzed.