Fault Detection Technology Research Articles

Unanticipated Fault Detection Technology for Tip/Tilt Mirror Control System of Large Aperture Telescope

Large Aperture telescopes are usually placed in some special environment. Once a malfunction occurs, it’s inconvenient for maintenance. This paper presents those unanticipated failures in tip/tilt mirror control system of large aperture telescopes, and tries to propose a detection strategy based on BP neural network. Firstly, the paper analyses the failure mechanism of the software and the hardware of the tip/tilt mirror control system. Then the known fault data are used to train the BP neural network in order to obtain an accurate detection model. Finally, the unanticipated data are mixed into the known fault data, and the model is used to detect unanticipated faults. The simulation results show that the model can be used to detect the unanticipated fault in tip/tilt mirror control system effectively.

Enhanced fault diagnosis method using conditional Gaussian network for dynamic processes

Applying fault detection and diagnosis (FDD) technology to the process industry can help to detect faults in time and minimize their impact. The purpose of this study is to propose an enhanced fault diagnosis method under a Conditional Gaussian Network(CGN) efficient and suitable for dynamic processes fault monitoring. The key paths are as follows: first, a time series model is established for the process data and decomposed into time-dependent components and time-independent components; second, time-dependent components are discarded and time-independent components void of auto-correlation are considered instead of the original data to learn the CGN model. A numerical simulation case is used to illustrate the interest of our proposal. The effectiveness of the proposed method is further verified and compared on the Tennessee Eastman Process (TEP). The obtained results show that our method has high and better accuracies regarding the diagnosis of known and unknown faults in dynamic processes.

Research on a Nonlinear Dynamic Incipient Fault Detection Method for Rolling Bearings

The incipient fault detection technology of rolling bearings is the key to ensure its normal operation and is of great significance for most industrial processes. However, the vibration signals of rolling bearings are a set of time series with non-linear and timing correlation, and weak incipient fault characteristics of rolling bearings bring about obstructions for the fault detection. This paper proposes a nonlinear dynamic incipient fault detection method for rolling bearings to solve these problems. The kernel function and the moving window algorithm are used to establish a non-linear dynamic model, and the real-time characteristics of the system are obtained. At the same time, the deep decomposition method is used to extract weak fault characteristics under the strong noise, and the incipient failures of rolling bearings are detected. Finally, the validity and feasibility of the scheme are verified by two simulation experiments. Experimental results show that the fault detection rate based on the proposed method is higher than 85% for incipient fault of rolling bearings, and the detection delay is almost zero. Compared with the detection performance of traditional methods, the proposed nonlinear dynamic incipient fault detection method is of better accuracy and applicability.

Open Fault Detection Technology of Bypass Circuit of PV Module with IR Camera

Open Fault Detection Technology of Bypass Circuit of PV Module with IR Camera

Analog circuit fault diagnosis via FOA-LSSVM

At present, the research on fault detection and diagnosis technology is very significant to improve the reliability of the equipment, which can greatly improve the safety and efficiency of the equipment. This paper proposes a new fault detection and diagnosis means based on the FOA-LSSVM algorithm. Experimental results demonstrate that the algorithm is effective for the detection and diagnosis of analog circuit faults. In addition, the model also demonstrate good generalization ability.

Data-driven process monitoring and fault analysis of reformer units in hydrogen plants: Industrial application and perspectives

Reformer boxes are complex, integrated, and high-temperature units, subject to various failures during continuous operations for extended time periods. Challenges in the development of high-fidelity first principle models, despite easy availability of process measurements motivated the development of data-driven, automated fault detection (FD) systems. Paucity of plant-wide implementation of FD technologies in the chemical industry, accentuates the absence of relevant practical guidelines and best practices. In this paper, a trivially replicable FD system has been developed for large-scale industrial reformer boxes of hydrogen manufacturing units. Actual process data from plant historian has been used for training and validation of a novel model, developed using a combination of partial least squares regression and principal components analysis. Abnormalities based on several important measurements around the reformer were identified. Explicit algorithmic details and insights obtained during development of the expert system have been provided for ease of replication and adaptability.

Optimized KLD-Based Fault Detection Method for Complex System Under Multi-Operation Conditions

With the increasing development of industrial technology, fault detection technology has been widely applied in many complex systems. Aiming at the problems of multi-operation conditions and the data length needed for fault detection using Kullback- Leibler Divergence (KLD), a novel fault detection method for complex systems based on optimized KLD under multi-operation conditions is proposed in this paper. In the first place, based on the historical data of the system, the complex operation conditions of the system are divided into several simple mutually exclusive operation conditions, and the division standard is established. In the next place, the optimal length can be quickly determined by autocorrelation analysis and applied to various operation conditions. After the next, in the training stage, the KLD values of all training data are calculated with the benchmark that is the initial optimal data length of each operation condition. And the maximum value is taken as the threshold for fault detection. Afterward, in the test phase, it starts by judging the type of operation condition that the data belongs to, then the corresponding KLD value is calculated, which is compared with the corresponding threshold, so as to determine whether the fault occurs. Eventually, this method is applied to the suspension system of the maglev train and respectively compared with the fault detection methods based on Euclidean distance or Mahalanobis distance. The results show that the proposed method possesses a low false alarm rate and high sensitivity.

Reviews of Research on Mechanical Fault Diagnosis in GIS

Gas Insulated Switchgear (GIS) has been widely used in recent years due to their advantages of small floor area, high reliability and good safety. It has arised a series of problems. At present, researchers at home and abroad have done more in-depth research on insulation fault, but research on mechanical fault is insufficient. Firstly, this paper reviews the research status of mechanical fault detection technology in GIS at home and abroad. Secondly, the aspect of the principles of mechanical fault in GIS, fault judgment and location technology, and early fault prediction and status evaluation in GIS were summarized. Finally, the paper prospected the future research direction on this basis.

Steady-State Process Fault Detection for Liquid Rocket Engines Based on Convolutional Auto-Encoder and One-Class Support Vector Machine

Liquid rocket engines (LREs) are the main propulsive devices of launch vehicles. Due to the complex structures and extreme working environments, LREs are also the components prone to failure. It is of great engineering significance to develop fault detection technologies which can detect fault symptoms in time and provide criteria for further fault diagnosis and control measures to avoid serious consequences during both the ground tests and flight missions. This paper presents a novel fault detection method based on convolutional auto-encoder (CAE) and one-class support vector machine (OCSVM) for the steady-state process of LREs. We train the CAEs by normal ground hot-fire test data of a certain type of large LRE for automatic feature extraction. Then the obtained features are used to train the OCSVMs to accomplish the fault detection task. The results demonstrate that the proposed method outperforms traditional redline system (RS), adaptive threshold algorithm (ATA), and back-propagation neural network (BPNN). We also study the effect of sample sizes and domain knowledge on the performance of the proposed method. The results suggest that appropriate measures that enrich the effective information content in the training data, such as increasing sample size and introducing domain knowledge, can further improve the performance of the proposed fault detection method.

Root causality analysis at early abnormal stage using principal component analysis and multivariate Granger causality

As fault detection technologies have been developed, process fault diagnosis at early abnormal stage has come to be considered a major problem. In this work, a method to analyze the root cause of faults is developed to provide proper information at the early abnormal stage. First, principal component analysis (PCA) is used for the early detection of the process fault. Then, the contributions, from which the normal portion is removed, are decomposed by singular value decomposition (SVD) method to select the hierarchical sensors. Finally, the multivariate Granger causality (MVGC) method is used to construct the sensor causalities using the hierarchical sensors. The developed methodology is verified using the liquefied natural gas fractionation process model, which embeds a sufficient number of highly correlated sensors. The results are compared with the conventional principal component analysis method and amplification of the residual contribution method to verify the advantages of the proposed method.

Health status diagnosis of distribution transformers based on big data mining

AbstractIn the fault detection technology of distribution transformer, the traditional artificial intelligence algorithm has been unable to achieve its efficient analysis and processing, but also unable to achieve the elimination of misdiagnosis caused by improper interval segmentation in the process of fault diagnosis of distribution transformer. In a word, the problem that always exists in transformer fault diagnosis technology is the problem of discretization of fault data. In order to solve this problem, this paper creatively proposes a health condition diagnosis method of distribution transformer based on large data. This method innovatively proposes that the dissolved gas analysis value of distribution transformer is the conditional attribute, and the fault type is the decision attribute, and the fault decision table is established. The continuous attribute data in the decision table are discretized by using the optimization behavior of large data sets. Subsequently, the discretized decision table is simplified by using the big data theory, and the decision table of fault diagnosis rules is established, which greatly simplifies the difficulty of attribute simplification of decision table and makes diagnosis more convenient. Finally, an example shows that the proposed method can effectively discretize and reduce samples. Compared with traditional methods, it improves the accuracy of fault diagnosis.

The research on Power Supply Array Intelligent Management Technology for Spacecraft Thermal Test

The thermal test of spacecraft has a large demand for the program-controlled power supply, which was configured as array. In this paper, the intelligent management technology of program-controlled power supply is researched. Base on the analysis of the history fault type of power supply, the fault detection technology of power supply is studied. Then, combined with the technology of power supply redundancy switching, an on-line switching device is designed. Through the verification experimental, the experimental results show that the system can quickly realize the detection and positioning of the fault power supply, and can finish the online switching of the fault power supply to the backup power within the second-level time. The research of intelligent management technology enhances the management efficiency of the large scale power supply array, and improves the operational reliability of the spacecraft thermal test.

Multi-block statistics local kernel principal component analysis algorithm and its application in nonlinear process fault detection

It is vital for fault detection technology to extract features of industrial process data effectively. Local kernel principal component analysis (LKPCA) has proved its good performance in preserving global and local structural characteristics. However, it ignored useful high-order statistics of data, so multi-block statistics local kernel principal component analysis (MSLKPCA) algorithm integrating statistics pattern analysis (SPA) into LKPCA is proposed. The correlation coefficient matrix is first calculated and K-means clustering is adopted to divide the original variables into several blocks. Then the weighted SPA, which gives different weights to different samples in each window according to their distributions, is adopted to build statistic spaces containing both low-order and high-order statistics. After that, LKCPA is performed in each statistic space to realize feature extraction. To reduce the noise effect amplified by SPA, PCA is adopted in the residual space to remove noise. Bayesian strategy is used to fuse the results of each block and two monitoring statistics ET and ER are proposed to monitor the feature space and the residual space respectively. The Tennessee–Eastman (TE) process simulation shows the effectiveness and superiority of the proposed algorithm for process monitoring.

Study on the Fault Detection Technology for Satellites In-Orbit

As the fault detection of satellites in-orbit concerned, characters of the telemetry data and conventional fault detection methods were analyzed . Based on this, the demand on the development of fault detection technologies was studied, and new methods were proposed with the application effects verified.

A Dilation and Erosion-Based Clustering Approach for Fault Diagnosis of Photovoltaic Arrays

Due to the incremental economic and environmental crisis, solar energy is becoming one of the most promising supplements for electric power generation. In order to eliminate the risks of safety hazards in the Photovoltaic system, fault detection technology is a priority research topic. Since the data distributions of different operation conditions present nonspherical clustering features, various supervised and semi-supervised intelligent learning methods have been widely carried out to detect and classify various fault conditions. However, the number of fault types needs to be determined in advance in these methods. This step takes a lot of time in the process of fault identification. In this paper, an original clustering method based on dilation and erosion theory is proposed. The main advantage of this approach is that the clustering number does not need to be predetermined. The simulation and experimental results verify that the proposed method has the high adaptability and validity under a wide range of meteorological conditions.

Development and implementation of automated fault detection and diagnostics for building systems: A review

Abstract This article reviews the current research on the development and implementation of automated fault detection and diagnostics (AFDD) technology for building systems. This article first examines the fundamentals of AFDD and its special requirements on building systems to provide a theoretical formulation of the problem. Some infrastructural barriers for scalable AFDD implementation are discussed. Then it reviews various methods used by previous researchers and real-life products in a typical AFDD workflow. The article compares different methods for feature generation and fault detection and fault diagnostics, and proposes ways to improve their current limitations from other research disciplines. The authors also discuss potential research topics to inspire further developed of new AFDD methodologies and make them more applicable, including: 1) fault assessments and improved information delivery; 2) benchmarking AFDD performance; 3) better interoperability of AFDD methods; and 4) cases of fault propagation and simultaneous faults.

Process Pattern-Based Near-Infrared Spectroscopy (NIRS) Fault Detection Using a Potential Function.

This paper proposes a near-infrared (NIR) fault detection technology based on a process pattern via a potential function. Near-infrared spectroscopy is used to acquire process information at the molecular level. In this study, the process pattern concept is first introduced in the field of process control and a process pattern construction method based on elastic net-PCA is put forth. Next, the potential function discriminant method is applied to distinguish and classify the constructed process pattern and identify the running state of the industrial system. Finally, the proposed method is verified and analyzed using spectra data of the crude oil desalination and dehydration process. Compared with existing fault detection methods, the proposed approach offers the following advantages: (1) potential function discrimination achieves nonlinear process classification with better fault detection accuracy and good visualization performance; (2) fault detection based on NIR spectra is faster with and possesses greater accuracy because it acquires process information from a microscopic molecular perspective; and (3) the process pattern contains more effective process information and can more comprehensively characterize the essential features of processes.

Open-Circuit Fault Analysis and Diagnosis for Indirect Matrix Converter

Due to the absence of bulky dc-link capacitors, the matrix converter (MC) is a prospective candidate to replace the conventional two-stage ac–ac converter, especially in some applications which require high reliability. In order to enhance the reliability during system operation, this paper investigates an open-circuit fault detection technology for indirect MC. First, combining with the characteristics of the space vector modulation technique, this paper presents analysis work during the appearance of an open-circuit fault. Then, a novel fault diagnosis approach is proposed based on the circuit analysis. The proposed method establishes a unique corresponding relationship between switches and modulation sectors, and the faulty switch can be discriminated exactly by measuring the three-phase output voltages and the dc-link voltage. This method, which can locate the failure that occurs in rectifier stage or inverter stage correctly, offers favorable antidisturbance ability. The experiments are carried out to verify the feasibility and validity of the proposed fault detection method.

On the Role of Minor Branches, Energy Dissipation, and Small Defects in the Transient Response of Transmission Mains

In the last decades several reliable technologies have been proposed for fault detection in water distribution networks (DNs), whereas there are some limitations for transmission mains (TMs). For TM inspection, the most common fault detection technologies are of inline types—with sensors inserted into the pipelines—and then more expensive with respect to those used in DNs. An alternative to in-line sensors is given by transient test-based techniques (TTBTs), where pressure waves are injected in pipes “to explore” them. On the basis of the results of some tests, this paper analyses the relevance of the system configuration, energy dissipation phenomena, and pipe material characteristics in the transient behavior of a real TM. With this aim, a numerical model has been progressively refined not only in terms of the governing equations but also by including a more and more realistic representation of the system layout and taking into account the actual functioning conditions. As a result, the unexpected role of the minor branches—i.e., pipes with a length smaller than the 1% of the length of the main pipe—is pointed out and a preliminary criterion for the system skeletonization is offered. Moreover, the importance of both unsteady friction and viscoelasticity is evaluated as well as the remarkable effects of small defects is highlighted.

High-impedance fault detection technology based on transient information in a resonant grounding system

High-impedance faults (HIFs) occur frequently in resonant grounding systems and are difficult to detect. Thus, an effective HIF feeder identification technology is urgently in demand. An HIF equivalent circuit in a resonant grounding system is established. The magnitude and phase relationships of transient zero-sequence currents among the faulty feeder, healthy feeders, detection points, and fault point, as well as the relationships between their projection coefficients with respect to the transient zero-sequence voltage are analysed. A novel HIF detection method based on the comparison of the magnitude and polarity of the transient zero-sequence current's projection coefficient is proposed. The accuracy of the proposed method is verified by simulations and field data.