Fault Threshold Research Articles

Co-Design of Observer-Based Fault Detection Filter and Dynamic Event-Triggered Controller for Wind Power System Under Dual Alterable DoS Attacks

In this article, a novel co- design approach of observed-based fault detection filter (FDF) and dynamic event-triggered controller is proposed for multi-area wind power system under dual alterable aperiodic (DAA) denial-of-service (DoS) attacks. It is the first attempt to design an observed-based FDF for wind power system considering actuator fault signal. And a tolerable actuator fault threshold function is constructed to warn the occurrence of fault signal. Then, considering the bandwidth limitation of network communication, a novel dynamic event-triggered scheme is proposed to decrease the occupation of communication channel. Furthermore, the DAA DoS attacks presented in this work have different blocking effects on data transmission in different network channel. On account of the existence of aperiodic DoS attacks, the general load frequency control model for wind power system is reconstructed as a new switched system. Based on the reconstructed model, the stability with an <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$H_{\infty }$ </tex-math></inline-formula> performance index is demonstrated by constructing appropriate polynomial Lyapunov-Krasovskii functionals. Besides, the FDF and dynamic event-triggered controller can be achieved by solving linear matrix inequalities. Finally, some contradistinctive case studies are given to illustrate the feasibility and effectiveness of the approaches proposed in this article.

An unsupervised prediction method for radar transmitter degradation fault based on isolation forest

The transmitter is prone to premature degradation and failure for the frequent work and the harsh working environment of the marine radars. The existing method of degradation fault based on data-driven require a large number of negative samples or need to set failure threshold. To solve the above problems, an unsupervised prediction method for radar transmitter degradation fault based on isolation forest is proposed. The proposed unsupervised prediction method can predict the degradation fault of radar transmitter when there are no fault samples and the fault threshold is not reached.

Energy fault detection for small buildings based on peer comparison of estimated operating status

As the penetration of smart meters grows, new energy conservation services using electricity consumption data have drawn much attention. Yet service frameworks for small buildings to provide alerts automatically when a failure in energy conservation is detected have not been sufficiently investigated. Thus, we propose a framework for detecting energy faults based on peer comparison of operating time estimated by self-organizing maps from smart meter data with a coarse time resolution. The framework first defines a common operating status for the business type using the results of operating time estimation from multiple buildings of the same type, and then monitors the operating status at 1-h intervals using only a fault threshold parameter specified by the user. If each estimated operating status is judged to be significantly larger than the average level for the building type, the system identifies it as an energy fault. The results of case studies using an open building energy management system dataset suggest that the proposed framework can detect failures in energy conservation caused by inappropriate HVAC management, which usually results in higher contract power consumption, or by failure to turn off appliances that unnecessarily consume energy after business hours.

Initial fault time estimation of rolling element bearing by backtracking strategy, improved VMD and infogram

Rolling bearing failure is widely regarded as a failure form of industrial machines. Owing to the poor operating circumstance with the stochastic contact between rolling elements, the performance of the bearing will deteriorate over time and cause a cascade breakdown in the mechanical system. Early fault detection has been found to be an effective strategy to avoid economic loss. Therefore, an integration method for fault diagnosis that combines backtracking strategy, improved variational mode decomposition (VMD), and infogram is proposed to tackle the challenge of the early feature extraction from the heavy noisy non-stationary signal. The backtracking strategy is adopted to track the data sample points earlier than the fault threshold determined based on the kurtosis index. The optimum parameters α and K of VMD are acquired through the particle swarm optimization (PSO) algorithm. In this way, the more accurate intrinsic mode functions (IMFs) can be gained by the improved VMD. The optimum IMFs are acquired according to the maximum values of kurtosis and correlation coefficients, and these IMFs can be reconstructed into the noise reduction signal. Since envelope analysis requires the selection of the appropriate central frequency and bandwidth, infogram is utilized to select the values of them. A simulated case is applied to demonstrate the validation of the proposed method. And to further illustrate its practicality, it is employed to perform early fault diagnosis for an experimental case. According to the diagnosis results, the proposed method has conspicuous superiority over the other existing technologies for estimating incipient fault time of the bearing.

Model-based strategy and surrogate function for health condition assessment of actuation devices

Prognostics and Health Monitoring (PHM) is a discipline aiming to determine in advance the Remaining Useful Life (RUL) of a system. To do so, the operation of the system is monitored in search of the early signs of degradation and incipient faults; then, a model for the propagation of faults is employed to estimate the propagation of damages and evaluate the RUL. Usually, a fault threshold is employed as a stopping criterion for the evaluation of damage propagation, but this is not a reliable method when dealing with multiple faults affecting the system at the same time. Specifically, the combined effect of two fault modes can cause the system not to meet its requirements well before the single faults reach their individual thresholds. In this work, we address a model-based strategy to estimate whether the system with incipient faults is still able to meet its performance requirements. The method is applied to aerospace actuators, and performance is evaluated in terms of dynamical response. This model-based algorithm is too slow to be evaluated in real-time, so a Support Vector Machine (SVM) is trained as a surrogate function to speed up the computation. The results and computational times of the full, physics based model and those of its surrogate are compared and discussed.

The Optimal Current Ratio Control of Redundant Electric Drive Systems and Diagnostic Strategies for Disagreement

As research into technology for autonomous vehicles becomes ever more intensive, the need for redundant system designs, as a solution for fail-operational functionality, has also increased. Redundant systems are those involved in maintaining normal operation even when one or more components fails. In particular, in electric power steering (EPS) systems, redundancy is achieved by applying dual-winding (DW), permanent magnet synchronous motors (PMSMs) and dual electronic control units (ECUs). In this study, the torque and efficiency of DW-PMSMs is analysed based on the manner in which the currents in each winding mechanism are combined. It was found that the PMSMs have maximum torque and efficiency when the current ratio was the same between DWs. A diagnostic strategy is proposed whereby the phase current ratio of dual lanes is not matched. The proposed diagnostic process includes the selection of the optimal fault threshold of the current difference between two ECUs and a method for distinguishing which of these was a failure for disagreement. In addition, depending on whether or not failure detection has occurred, a current compensation control method is proposed for reverting to normal operation or for failure operations. The dual-lane system was fabricated and the proposed solutions were experimentally verified.

Online Short-Circuit Protection Strategy of an Electric Powerpack for Electric Oil Pump Applications

With the trend towards the miniaturisation and electrification of automotive components, the electric oil pump (EOP) is generally designed as an integrated powerpack in which a motor and an electronic control unit (ECU) are combined into an integrated system, with the electronic components of the ECU exposed to harsher temperature environments. The critical weakness of the traditional design is short-circuit failures, and the most common protection method is $V_{ds}$ monitoring. However, this method cannot predict the accurate protection current, indicating that the fault detection level and short-circuit current (SCC) may change depending on the operating temperature and load condition to which the devices are subject. In this case, switching devices can be destroyed or deteriorated if they are unexpectedly placed outside of the protection range. This study proposes an online short-circuit protection method in which the optimal fault threshold can be controlled in real time, in accordance with changes in temperature and load conditions. To this end, this study investigates the cause of fluctuations in SCC and analyses the protection range and qualification time for SCC through worst-case analysis. To determine the optimal fault threshold, the SCC waveform is estimated through circuit analysis and a method of estimating $R_{dson}(T_{j})$ , which is the cause of SCC fluctuation, is proposed based on simulation and experimental data. Finally, according to $R_{dson}(T_{j})$ , optimal thresholds are derived and the proposed methods are experimentally-verified. It is found that the SCC could be predicted and managed within the protection range by the proposed new protection method.

Research on battery pack fault pre diagnosis method

At present, most of the BMS in the market set a constant fault threshold, but with the increase of the number of power battery recycling, the external characteristics of the battery will also change. A constant fault diagnosis threshold may also occur after battery aging diagnosis is not timely. Battery cell voltage is the direct embodiment of the external characteristics of battery cell, so the fault diagnosis of battery cell voltage is the most important. Too high or too low cell voltage is the most obvious fault. Normally, this should not happen. Of course, due to the gradual aging of the battery pack, the inconsistency between the monomers and the charged state of each cell, the inconsistency of the voltage of each cell can not be avoided. However, the voltage difference caused by cell manufacture or inconsistency is limited in the case of aging degree.Based on the battery discharge data, the purpose of this study is to propose a method to identify the gradual fault signal in advance and reduce the problems caused by the sudden failure of the battery by detecting and diagnosing the change trend of the extreme value difference of the historical operation data.

Substation DC system grounding fault prediction method

There may be disturbance and uncertainty in the collection of leakage current in DC system of substation, which leads to the decrease of accuracy and increase of prediction error. Based on this, an improved grey prediction method is proposed to predict DC system branch grounding fault. Firstly, the characteristics of DC system ground fault parameters are collected. Secondly, the improved grey prediction algorithm is used to predict and estimate whether the detection reaches the fault threshold in the future. Finally, the validity of the proposed method is verified by MATLAB modeling.

Model-Based Analysis and Quantification of Bearing Faults in Induction Machines

The detection of rolling-element bearing fault can be accomplished by monitoring and interpreting a variety of signals, including the vibration, the acoustic noise, and the stator current. The existence of a bearing fault as well as its specific fault type can be readily determined by performing frequency spectral analysis on the monitored signals with various signal processing techniques. However, this traditional approach, despite being simple and intuitive, is not able to identify the severity of a bearing fault in a quantitative manner. Moreover, it is often times tedious and time-consuming to apply this approach to electric machines with different power ratings, as the bearing fault threshold values need to be manually calibrated for each motor running at every possible speed and carrying any possible load. This article, thus, proposes a quantitative approach to estimate the bearing fault severity based on the airgap displacement profile, which is reconstructed from the mutual inductance variation profile estimated from a quantitative electrical model that takes the stator current as input. In addition, the accuracy of the developed electrical model and the estimated bearing fault severity are validated by the simulation and experimental results, and the explicit airgap variation profile is reconstructed with the superposition of multiple Fourier series terms estimated from the stator current via the proposed scheme. The proposed method offers a quantitative and universal bearing fault indicator for induction machines with any power ratings and operating under any speed and load conditions.

Fault Detection for Suspension System of Maglev Trains Based on Historical Health Data

To overcome the influence of multiple operating conditions for fault detection, this paper proposes a method to detect fault for the suspension system of maglev trains. Firstly, the complex operating condition of the maglev train is divided into some simple conditions, and the operating samples are extracted through the time window in the same simple operating condition. Secondly, the features of the extracted samples are extracted by the Fast Walsh-Hadamard transform, and the noise is removed by the median filtering. Thirdly, after adopting principal component analysis to reduce the dimensionality of the feature, the mean of the feature is calculated and applied to calculate the Euclidean distance from the features of the new data obtained by the same processing in each time. Fourthly, the new Euclidean distance obeying the Gaussian distribution is obtained through the Box-Cox transformation. Finally, the fault threshold in each simple operating condition is established through the characteristics of Gaussian distribution. Taking a certain operating condition as an example, the method is compared with the three similar methods. The results reveal that the proposed method can detect the faults timely and has a low false alarm rate. Moreover, the proposed sub-healthy data are within an acceptable range. Based on these results, the proposed method can be applied in fault detection of the maglev train, providing a certain basis for fault diagnosis.

Research on Fault Diagnosis of Wind Turbine Based on SCADA Data

Effective early warning of wind turbine failures is of great significance to reduce the operation and maintenance costs of wind farms and improve power generation efficiency. At present, most wind farms are installed with supervisory control and data acquisition (SCADA) system, and SCADA data contains a lot of hidden information, which can be used for fault early warning. This paper uses the generator temperature and gearbox oil temperature in the SCADA data as the entry point for fault warning. Firstly, the eXtreme gradient boosting (XGBoost) algorithm is used to establish the normal temperature regression prediction model of wind turbine components. Then, the residual between the predicted value and the actual value is calculated, and the change trend of the residual is monitored by the principle of exponentially weighted moving-average (EWMA) control chart. Finally, by setting an appropriate threshold, the variation trend of the residual is judged to determine the occurrence and development of the fault. This paper uses two fault detection methods: fixed threshold and dynamic threshold based on adaptive algorithm, and compares the advantages and disadvantages of the two methods. Based on the SCADA data of a wind farm in Inner Mongolia (China), this paper designs the fault early warning test of the wind turbine generator and gearbox. The experimental results show that for the generator, the fixed fault threshold method can give the fault alarm 3 hours in advance, while the dynamic fault threshold determination method can give fault alarm 4.25 hours in advance. For gearbox, the fixed fault threshold method can give the fault alarm 2 hours in advance, while the dynamic threshold fault diagnosis method can send out the fault alarm 2.75 hours in advance.

An Effective Stator Fault Diagnosis Framework of BLDC Motor Based on Vibration and Current Signals

Electric motor is a prominent rotary machinery in many engineering applications due to its excellent electrical energy utilization. With the increased demand in production and complex operating conditions, motors often run in a severe loading condition. Overload, overheating and many other intricate operating conditions account for the stator related faults in motors. Motor current signature analysis (MCSA) and vibration analysis have been popular techniques to diagnose different stator and rotor related faults in motors. But it is difficult to find the fault magnitude or fault threshold by using only one approach due to nonstationary motor operations. This paper presents a comprehensive review of a permanent magnet brushless DC motor’s (BLDC motor) fault diagnosis combining vibration and current signals collected from sensors. Since the insulation break in the stator winding is the most commonly occurring fault in the stator, a short-circuit was artificially created between two windings. Based on the motor operating conditions, three health states are chosen from the experimental sensor data with different fault magnitudes. Health states are labeled as healthy state, incipient failure state, and severe failure state. Two effective fault diagnosis indices named kurtosis and third harmonic of motor current are selected for analyzing the vibration signals and current signals, respectively. Proposed diagnostics framework is validated using experimental data and proven to detect the stator fault at the early stage as well as distinguish between different fault states. Monitoring both mechanical and electrical characteristics of BLDC motor provides a thorough understanding of fault magnitude and threshold in different health states.

A Degradation Fault Prognostic Method of Radar Transmitter Combining Multivariate Long Short-Term Memory Network and Multivariate Gaussian Distribution

In the prognosis of radar transmitter degradation fault, there are some problems, such as the total sample size and fault sample size of sensor monitoring data are small, and the monitoring data can not reach the fault threshold. To solve these problems, a prediction model combining the multivariate long short-term memory networks with multivariate Gaussian distribution is proposed, in which the long short-term memory networks predict the subsequent time step of multi-sensor monitoring data, and the multivariate Gaussian distribution model constructed by few fault samples is used to realize the prediction of degraded faults. The key parameters of the model are determined by the data processing experiments of double monitoring points, and the validity of the model is verified by the data processing experiments of multiple monitoring points. The experimental results show that the degradation fault can be predicted effectively within 10% of the total time series of monitoring data. Compared with the traditional radar warning after failure, the model can effectively predict the degradation fault of the transmitter when the fault sample size is low and the fault threshold is not reached.

A novel data-temporal attention network based strategy for fault diagnosis of chiller sensors

In the air-cooled chiller system, sensor fault diagnosis has great significance for ensuring normal operation. However, according to the operating mechanism of the chiller system, sensors’ readings exhibit dynamical data-temporal dependencies and are easily affected by external factors and control parameters. To further capture the data characteristics existed in the sensors time series, in this paper, we propose a novel data-temporal attention network (DAN) for the chiller sensor fault diagnosis. Based on the conventional encoder-decoder network (EDN), the proposed DAN model is firstly built by adding three novel parts: the first one is a data attention mechanism embedded in the encoder, which is used to capture the dynamic data correlation between different sensors; the second part is a temporal attention mechanism, which is employed in the decoder to model the dynamic time-dependencies among the sensors time series; considering the influence of external factors and control parameters, the third part is a fusion module to incorporate these influential factors from different domains. Thereafter, we design a specific chiller sensor fault diagnosis strategy using the proposed DAN model. The sensor fault diagnosis strategy uses only normal sequences for training and learns to reconstruct normal time series behaviors, and then determines the fault threshold of each chiller sensor, and finally identifies the specific sensor fault by comparing the absolute reconstruction error vector with the fault threshold vector. In the end, the experiments which adopt data sets from a real air-cooled chiller platform are conducted, and detailed comparisons are made. Various magnitudes of fixed biases are introduced into eleven sensors for validation. Experimental results reveal that the sensor fault diagnosis strategy with the proposed DAN model achieves the best training and fault diagnosis performance compared with its variants and the traditional EDN model. Especially for the sensor fault diagnosis performance, comparison results demonstrate that the proposed DAN model is more sensitive to the small biases than the other contrast models and has better robustness for impacts on fault sensor in the reconstruction of the fault-free sensors.

Multi-Model Switching Based Fault Detection for the Suspension System of Maglev Train

In order to satisfy the requirements of on-line monitoring of the suspension system under complex conditions, a fault detection method for maglev train suspension system based on multi-model switching is proposed. In the proposed method, the healthy samples are extracted through the moving time window, and then the features of the healthy samples are extracted by the Fast Walsh–Hadamard transform and filtered by the median filter. Then, the normalization is used to eliminate the difference of feature vectors, and then the principal component analysis method is used to reduce the dimension and de-correlation of the feature matrix contributing to a hyper-sphere space. Finally, the health threshold and the fault threshold are determined by Euclidean distance, then fault detection models are established for various operating conditions. Taking the positive line operation condition as an example, the proposed method is compared with the method based on the original feature and support vector data description based on the original feature. The results demonstrate that the proposed method is superior to the other two methods in terms of health detection rate and false positive rate. In addition, the proposed method is of the characteristics of low computational complexity, no-parameter optimization, and good robustness. It can be applied in practical engineering providing a certain research basis for data deep mining such as fault diagnosis and fault prediction.

Research on Residual Life Prediction for Electrical Connectors Based on Intermittent Failure and Hidden Semi-Markov Model

Based on the dynamic properties of electrical connector intermittent failure, the model and methods for residual life prediction for electrical connectors are studied in this paper. Firstly, the mechanism of electrical connector intermittent failure is analyzed, and the area enclosed by the contact resistance curve and the fault threshold is defined as the generalized severity of intermittent failure to describe how severe the electrical connector’s intermittent failure is. Then, the Hidden Semi-Markov Model (HSMM) is introduced to build the residual life prediction model of the electrical connector. Further, the evaluation method of using the state and prediction method for residual life are studied. Finally, by carrying out the residual life prediction test, the effectiveness of the residual life prediction method for electrical connectors based on intermittent failure and HSMM is verified.

An On-Line Condition Monitoring System for Incipient Fault Detection in Double-Cage Induction Motor

This paper presents a complete on-line condition monitoring system (CMS) designed to detect incipient broken rotor bar (BRB) fault in double-cage induction motor (DCIM) using the stator current signature. The novelty of this paper lies in the successful combination of one of the latest variants of wavelet techniques, the recursive stationary wavelet packet transform and the widely used tool in quality control, the statistical process control (SPC) in order to deal with several challenges in the continuous monitoring of incipient BRB fault. In fact, breakage in outer bars (OBs) of a DCIM is known to be less prone to exhibit fault signature in the stator current than that of a single-cage IM, especially during the steady motor operation mode. Furthermore, on-line CMSs should at the same time act at normal motor operating mode, including steady and transient modes and be reconfigurable with no motor size dependence. The complete on-line CMS was designed and implemented in a low-cost hardware environment and was validated with 3-kW DCIM featuring incipient OB faults. A 100% fault detection accuracy starting from 3-mm OB breakage during normal motor operating mode, with the autolearning fault threshold, demonstrates the reliability and flexibility of the proposed system.

Morphological Fault Detector for Adaptive Overcurrent Protection in Distribution Networks With Increasing Photovoltaic Penetration

In this paper, a versatile and robust algorithm for overcurrent protection that is adaptive to the feeder PV penetration level as well as changing load condition in the distribution feeder is proposed. The algorithm has built-in dc-offset suppression capability as well as recursive least square error filter for current phasor estimation to provide input to the overcurrent fault detector. The proposed multistage morphological fault detector analyzes sudden change in fault current magnitude and its structure makes it convenient to update the fault threshold dynamically by measuring the current magnitude at regular interval, thus making the algorithm versatile and adaptive. The performance of the proposed algorithm is investigated by simulating different fault distances and fault types for increasing penetration levels of PVs on a simple radial distribution feeder .

DNN‐based approach for fault detection in a direct drive wind turbine

Incipient fault detection of wind turbines is beneficial for making maintenance strategy and avoiding catastrophic result in a wind farm. A deep neural network (DNN)-based approach is proposed to deal with the challenging task for a direct drive wind turbine, involving four steps: a preprocessing method considering operational mechanism is presented to get rid of the outliers in supervisory control and data acquisition (SCADA); the conventional random forest method is used to evaluate the importance of variables related to the target variable; the historical healthy SCADA data excluding outliers is used to train a deep neural network; and the exponentially weighted moving average control chart is adopted to determine the fault threshold. With the online data being input into the trained deep neural network model of a wind turbine with healthy state, the testing error is regarded as the metric of fault alarm of the wind turbine. The proposed approach is successfully applied to the fault detection of the fall off of permanent magnets in a direct drive wind turbine generator.