Online Fault Detection Research Articles

Fault detection and diagnosis in multivariate systems using multiple correlation regression

Correlations among variables are inherent features of a multivariate system. Making a good use of variable correlations is important to fault detection and diagnosis (FDD). In this paper, a novel FDD method is proposed using multiple correlation regression (MCR). MCR builds a regression model using the multiple correlation of a variable with other variables. MCR is the best linear regression that produces residuals with the minimal variance. Based on the MCR residuals of variables, a T2 statistic is defined for fault detection. This T2 statistic consists of a set of independent components, with each component corresponding to the MCR residual of a variable. To reduce the amplification and masking effects caused by fault-free components, three improvements to the residual-based T2 statistic are proposed, called the intersecting T2 (IT), weighted T2 (WT) and enhanced T2 (ET) statistics. In particular, the WT and ET statistics and their control limits vary with samples to adapt to dynamic changes of the system. An online fault detection strategy is proposed by combining the WT and ET statistics. The WT statistic is firstly used to detect the occurrence of faults. After faults were detected, the ET statistic is used to replace the WT statistic in order to enhance the capability of detecting subsequent faulty samples. A contribution-based fault diagnosis method is also developed. Faulty variables are identified by comparing variable contributions to the key faulty component of the residual-based T2 statistic. The implementation and advantages of the proposed FDD method are illustrated with a simulation example and an industrial case study.

On-line Detection and Classification of PMSM Stator Winding Faults Based on Stator Current Symmetrical Components Analysis and the KNN Algorithm

The significant advantages of permanent magnet synchronous motors, such as very good dynamic properties, high efficiency and power density, have led to their frequent use in many drive systems today. However, like other types of electric motors, they are exposed to various types of faults, including stator winding faults. Stator winding faults are mainly inter-turn short circuits and are among the most common faults in electric motors. In this paper, the possibility of using the spectral analysis of symmetrical current components to extract fault symptoms and the machine-learning-based K-Nearest Neighbors (KNN) algorithm for the detection and classification of the PMSM stator winding fault is presented. The impact of the key parameters of this classifier on the effectiveness of stator winding fault detection and classification is presented and discussed in detail, which has not been researched in the literature so far. The proposed solution was verified experimentally using a 2.5 kW PMSM, the construction of which was specially prepared for carrying out controlled inter-turn short circuits.

Soft Sensing for On-Line Fault Detection of Ammonium Sensors in Water Resource Recovery Facilities.

The increasing demand for online sensors applied to advanced control strategies in water resource recovery facilities has resulted in the increasing investigation of fault-detection methods to improve the reliability of sensors installed in harsh environments. The study herein focuses on the fault detection of ammonium sensors, especially for effluent monitoring, given their potential in ammonium-based aeration control applications. An artificial neural network model was built to predict the ammonium content in the effluent by employing the information from five other sensors installed in the activated sludge tank: NH4+, pH, ORP, DO, and TSS. The residual between the model prediction and the effluent ammonium sensor signal was utilized in a fault-detection mechanism based on principal component analysis and Shewhart monitoring charts. In contrast to previous studies, the present work utilizes typical faults collected from a 1 year historic dataset of an actual sensor setup. Treatment process anomalies, calibration bias faults, and fouling drifts were the most common issues identified from the historic dataset, and they were promptly identified by the proposed fault-detection methodology. Once a fault is detected, the model prediction can be actively used in place of the sensor for process control without affecting the treatment process by utilizing faulty datasets.

On the application of machine learning for defect detection in L-PBF additive manufacturing

This paper investigates the performance of several Machine Learning (ML) techniques for online defect detection in the Laser Powder Bed Fusion (L- PBF) process. The research aims to improve the consistency in product quality and process reliability. The applications of acoustic emission (AE) sensor to receive elastic waves during the printing process is a cost-effective way of materializing such a demand. In this study, the process parameters were intentionally adjusted to create three different levels of defects in H13 tool steel samples. The first class was printed with minimum defects, the second class had only intentional cracks, and the last class included both intentional cracks and porosities. The AE signals were acquired during the samples' manufacturing, and three different machine learning (ML) techniques were applied to analyze and interpret the dataset. First, a hierarchical K-means clustering is employed for labeling the data, followed by a supervised deep learning neural network (DL) to match acoustic signal with defect type. Second, a principal component analysis (PCA) was used to reduce the dimensionality of the dataset. A Gaussian Mixture Model (GMM) was then employed to enable fast defect detection suitable for online monitoring. Third, a variational auto-encoder (VAE) approach is used to obtain a general feature of the signal that could be an input for the classifier. A supervised DL trained on the H13 tool steel dataset successfully detected quality trends in AE signals collected from 316L samples. The VAE approach presents a novel method to detect defects in L-PBF processes that eliminate the need for model training in different materials.

Online defect detection and automatic grading of carrots using computer vision combined with deep learning methods

The demand for smart automatic system in postharvest technology, particularly in the postharvest of carrot production is high. In this paper, an automatic carrot grading system was developed based on computer vision and deep learning, which can automatically inspect surface quality of carrots and grade washed carrots. Specifically, based on ShuffleNet and transfer learning, a lightweight deep learning model (CDDNet) was constructed to detect surface defects of carrots. Carrot grading methods were also proposed based on minimum bounding rectangle (MBR) fitting and convex polygon approximation. Experimental results showed that the detection accuracy of the proposed CDDNet was 99.82% for binary classification (normal and defective) and 93.01% for multi-class classification (normal, bad spot, abnormity, fibrous root), and demonstrated good performance both in time efficiency and detection accuracy. The grading accuracy of MBR fitting and convex polygon approximation was 92.8% and 95.1% respectively. This research provides a practical method for online defect detection and carrot grading, and has great application potential in commercial packing lines.

Fault detection based on ICA-GLR for non-Gaussian industrial processes

ABSTRACT As the growth of Industry 4.0, online fault detection plays a crucial role in ensuring the manufacturing quality. Generally, the fault detection methods can be classified into model-based and data-driven methods. There are advantages/disadvantages between two methods. In this study, we integrated both methods in order to develop an efficient fault detection method for non-Gaussian industrial processes. The data-driven method, independent component analysis (ICA) is used to extract non-Gaussian information and dimensionality reduction. Meanwhile, the model-based method, generalized likelihood ratio (GLR) test is adopted as the charting statistic. The proposed ICA-GLR method has advantages of 1) detecting a wide range of process changes, 2) estimating the change points and 3) needless prior parameters to be specified by practitioner. The efficiency of the proposed ICA-GLR fault detection method will be verified via implementing one simulated non-Gaussian process and two real manufacturing processes: Tennessee Eastman process and semiconductor manufacturing process. Results demonstrate that the proposed ICA-GLR method has superior fault detectability when compared to traditional methods, such as principal component analysis and ICA.

Cost-Sensitive Extremely Randomized Trees Algorithm for Online Fault Detection of Wind Turbine Generators

The number of normal samples of wind turbine generators is much larger than the number of fault samples. To solve the problem of imbalanced classification in wind turbine generator fault detection, a cost-sensitive extremely randomized trees (CS-ERT) algorithm is proposed in this paper, in which the cost-sensitive learning method is introduced into an extremely randomized trees (ERT) algorithm. Based on the classification misclassification cost and class distribution, the misclassification cost gain (MCG) is proposed as the score measure of the CS-ERT model growth process to improve the classification accuracy of minority classes. The Hilbert-Schmidt independence criterion lasso (HSICLasso) feature selection method is used to select strongly correlated non-redundant features of doubly-fed wind turbine generators. The effectiveness of the method was verified by experiments on four different failure datasets of wind turbine generators. The experiment results show that average missing detection rate, average misclassification cost and gMean of the improved algorithm better than those of the ERT algorithm. In addition, compared with the CSForest, AdaCost and MetaCost methods, the proposed method has better real-time fault detection performance.

A novel corrosion detection method using wavelet transformed mode shapes of a functionally graded rotor-bearing system

Development of online fault detection based on vibration characteristics of a rotating system has been the subject of interest for the last few decades. A novel corrosion detection method is suggested in the present work, based on the wavelet transformed modal analysis of a functionally graded (FG) rotor system. Material properties of an FG shaft are position and temperature dependant and graded in a radial direction following the power-law and non-linear temperature distribution (NLTD) respectively. An FG rotor consisting of a mixture of Zirconium Dioxide (ZrO2) and Stainless Steel (SS) is considered, where the volume fraction of metal decreases toward the outer radius and ceramic volume fraction increases. The finite element (FE) formulations have been developed using Timoshenko beam theory (TBT) for modal analysis of a corroded FG rotor bearing system. A spatial variation of a wavelet transform is used to transform the computed FG rotor mode shapes into the wavelet domain to identify and locate the local corrosion defect. A sensitivity analysis and a parametric study has been carried out to further verify the validity of the suggested novel method.

Design and development of Cu‐Al‐Mn‐Ni shape memory alloy coated optical fibre sensor for condition‐based monitoring of physical systems

Online fault detection, isolation and recovery using smart sensors play an important role in intelligent manufacturing system. Fibre optic sensors are very interesting for condition monitoring applications due to the advantage of this technology. Here, the experimental demonstration of Cu-based shape memory alloy (SMA) coated optical fibre for temperature-based sensing applications is reported. The benefit of Cu-based SMA coated optical fibre over conventional metallic coating has been evaluated in the study. For consistent coating, an in situ fixture with a rotary drive setup has been designed and developed. Thermo optic test bench has been developed to study the actuation characteristics of the SMA coated optical fibre for varying current and voltage. Experiments were performed to investigate the light intensity in the SMA coated optical fibre at different actuation conditions. The displacement that takes place in the optical fibre due to the external temperature stimuli will create proportional intensity and wavelength shifts. The maximum average displacement of 4.9 mm has been achieved for Cu-Al-Mn-Ni coated optical fibre. Results show variation in the optical fibre signal due to heating and cooling of the fibre from the applied electrical stimulus on Cu-based SMA coating in the form of austenite to martensite transformation.

Automated fault detection for additive manufacturing using vibration sensors

ABSTRACT Online process control is a crucial task in modern production systems that use digital twin technology. The data acquisition from machines must provide reliable and on-the-fly data, reflecting the exact status of the ongoing process. This work presents an architecture to acquire data for an Additive Manufacturing (3D printer) process, using a set of consolidated Internet of Things (IoT) technologies to collect, verify and store these data in a trustful and secure way. The need for online monitoring and fault detection is addressed by the development of a classifier using Convolutional Neural Networks. This deep learning approach, using temporally aligned vibration data provided by the underlying architecture, allows raw data processing to detect patterns without signal pre-processing and without domain-specific knowledge for model building.

A Framework for Monitoring and Fault Diagnosis in Nuclear Power Plants Based on Signed Directed Graph Methods

When nuclear power plants (NPPs) are in a state of failure, they may release radioactive material into the environment. The safety of NPPs must thus be maintained at a high standard. Online monitoring and fault detection and diagnosis (FDD) are important in helping NPP operators understand the state of the system and provide online guidance in a timely manner. Here, to mitigate the shortcomings of process monitoring in NPPs, five-level threshold, qualitative trend analysis (QTA), and signed directed graph (SDG) inference are combined to improve the veracity and sensitivity of process monitoring and FDD. First, a three-level threshold is used for process monitoring to ensure the accuracy of an alarm signal, and candidate faults are determined based on SDG backward inference from the alarm parameters. According to the candidate faults, SDG forward inference is applied to obtain candidate parameters. Second, a five-level threshold and QTA are combined to determine the qualitative trend of candidate parameters to be utilized for FDD. Finally, real faults are identified by SDG forward inference on the basis of alarm parameters and the qualitative trend of candidate parameters. To verify the validity of the method, we have conducted simulation experiments, which comprise loss of coolant accident, steam generator tube rupture, loss of feed water, main steam line break, and station black-out. This case study shows that the proposed method is superior to the conventional SDG method and can diagnose faults more quickly and accurately.

Adaptive Multi-View Image Mosaic Method for Conveyor Belt Surface Fault Online Detection

In order to improve the accuracy and real-time of image mosaic, realize the multi-view conveyor belt surface fault online detection, and solve the problem of longitudinal tear of conveyor belt, we in this paper propose an adaptive multi-view image mosaic (AMIM) method based on the combination of grayscale and feature. Firstly, the overlapping region of two adjacent images is preliminarily estimated by establishing the overlapping region estimation model, and then the grayscale-based method is used to register the overlapping region. Secondly, the image of interest (IOI) detection algorithm is used to divide the IOI and the non-IOI. Thirdly, only for the IOI, the feature-based partition and block registration method is used to register the images more accurately, the overlapping region is adaptively segmented, the speeded up robust features (SURF) algorithm is used to extract the feature points, and the random sample consensus (RANSAC) algorithm is used to achieve accurate registration. Finally, the improved weighted smoothing algorithm is used to fuse the two adjacent images. The experimental results showed that the registration rate reached 97.67%, and the average time of stitching was less than 500 ms. This method is accurate and fast, and is suitable for conveyor belt surface fault online detection.

Intelligent Fault Detection, Diagnosis and Health Evaluation for Industrial Robots

The focus of this study is development of an intelligent fault detection, diagnosis and health evaluation system for real industrial robots. The system uses principal component analysis based statistical process control with Nelson rules for online fault detection. Several suitable Nelson rules are chosen for sensitive detection. When a variation is detected, the system performs a diagnostic operation to acquire features of the time domain and the frequency domain from the motor encoder, motor current sensor and external accelerometer for fault diagnosis with a multi-class support vector machine. Additionally, a fuzzy logic based robot health index generator is proposed for evaluating the health of the robot, and the generator is an original design to reflect the health status of the robot. Finally, several real aging-related faults are implemented on a six-axis industrial robot, DRV90L7A6213N by Delta Electronics, and the proposed system is validated effectively by the experimental results.

Low-Cost Embedded Real-Time Handheld Vibration Smart Sensor for Industrial Equipment Onsite Defect Detection

In recent years, many industrialfacilities have been required to install sensing system to monitor the health status of the equipment. Although the PC-based diagnosis system is a good alternative, it undoubtedly increases the hardware cost greatly, and also causes restrictions on system assembly. In some critical environments, installation of the PC-based system is not suitable because of high temperature, dust, and humidity. On the contrary, a high mobility portable smart sensor will be relatively adequate for onsite inspections. To solve this problem, we developed a low-cost embedded handheld smart sensor. Instead of using sophisticated algorithms, certain frequently used statistic indicators are considered. Nevertheless, due to the resource limitations of the embedded systems, it causes difficulty for real-time realization and therefore, the recursive architecture of the statistic indicators is derived. These statistical factors are then fed into a Gaussian classifier for online defect detection, which gives a great contribution to field operators for onsite inspections. The highly integrated hardware/software co-design of the developed device provides user-friendly and high mobility for field inspections. Finally, a practical industrial application regarding the online diagnosis of a solenoid valve actuator defect is presented to demonstrate the effectiveness of the developed embedded smart sensor.

Industrial Printing Image Defect Detection Using Multi‐Edge Feature Fusion Algorithm

Online defect detection system is a necessary technical measure and important means for large‐scale industrial printing production. It is effective to reduce artificial detection fatigue and improve the accuracy and stability of industry printing line. However, the existing defect detection algorithms are mainly developed based on high‐quality database and it is difficult to detect the defects on low‐quality printing images. In this paper, we propose a new multi‐edge feature fusion algorithm which is effective in solving this problem. Firstly, according to the characteristics of sheet‐fed printing system, a new printing image database is established; compared with the existing databases, it has larger translation, deformation, and uneven illumination variation. These interferences make defect detection become more challenging. Then, SIFT feature is employed to register the database. In order to reduce the number of false detections which are caused by the position, deformation, and brightness deviation between the detected image and reference image, multi‐edge feature fusion algorithm is proposed to overcome the effects of these disturbances. Lastly, the experimental results of mAP (92.65%) and recall (96.29%) verify the effectiveness of the proposed method which can effectively detect defects in low‐quality printing database. The proposed research results can improve the adaptability of visual inspection system on a variety of different printing platforms. It is better to control the printing process and further reduce the number of operators.

Induction Machine Stator Fault Tracking Using the Growing Curvilinear Component Analysis

Detection of stator-based faults in Induction Machines (IMs) can be carried out in numerous ways. In particular, the shorted turns in stator windings of IM are among the most common faults in the industry. As a matter of fact, most IMs come with pre-installed current sensors for the purpose of control and protection. At this aim, using only the stator current for fault detection has become a recent trend nowadays as it is much cheaper than installing additional sensors. The three-phase stator current signatures have been used in this study to observe the effect of stator inter-turn fault with respect to the healthy condition of the IM. The pre-processing of the healthy and faulty current signatures has been done via the in-built DSP module of dSPACE after which, these current signatures are passed into the MATLAB® software for further analysis using AI techniques. The authors present a Growing Curvilinear Component Analysis (GCCA) neural network that is capable of detecting and follow the evolution of the stator fault using the stator current signature, making online fault detection possible. For this purpose, a topological manifold analysis is carried out to study the fault evolution, which is a fundamental step for calibrating the GCCA neural network. The effectiveness of the proposed method has been verified experimentally.

On-line Fault Detection of Transmission Line Using Artificial Neural Network.(Dept.E)

As the voltage and current waveforms arc deformed due to transient during faults, their patient changes according to the type of fault. The Artificial Neural Network (ANN) c2 then be used for fault detection due to its distinguished behavior in pattern recognition. In order to minimize the structure and timing of the ANN, preprocessing of the voltage and current waveforms was done, The data delivered from a simulated power system using PSCAD (EMTP with cad system) was used for training and testing the ANN. An experimental setup, consists of a 3 phase power supply module and transmission line module, is utilized. A set of signal conditioning circuits is designed and implemented in order to transfer data to a PC which is used as an on-line relay for fault detection. This is done via a data acquisition card (CIO DAS1602/12). The Matlab program captures and processes real data for training the ANN. Applying different types of saults for testing the system, right tripping action was taken and the type of fault was correctly identified. The suggested artificial neural network algorithm has been found simple and effective hence could be implemented in practical application.

An Improved k-Value Symmetrical Difference Analytic Energy Operator With HTFIF and L-KCA for Bearing Fault Diagnosis

The opportune fault detection of the rolling element bearings can avoid serious equipment accidents or even casualties. However, the early fault features of the bearings are weak and often submerged in heavy background noise and interferences. This paper proposes a novel approach for bearing fault diagnosis based on the hard thresholding fast iterative filtering (HTFIF), IMF selection index integrating L- kurtosis, correlation coefficient and autocorrelation function impulse harmonic to noise ratio (L-KCA), and improved k-value symmetrical difference analytic energy operator (k-SDAEO). As an adaptive and fast time-frequency analysis method, HTFIF is first adopted to process the bearing vibration signal and obtain a series of IMFs. After that, an alternative fusion index L-KCA is developed to select the sensitive IMF. Finally, a novel k-SDAEO with strong noise robustness is presented to process the selected IMF. With this method, the weak bearing fault signatures can be identified from the energy spectrum. The performance of the proposed method comparing to the traditional methods are investigated by numerical simulation and experimental studies. The results show that the proposed method has better fault feature extraction capability and higher efficiency, which can be implemented in the on-line and real-time fault detection.

Developing A Supervisory Technique for Online Fault Isolation and Coordination Tasks (Dept.E)

The control of complex automated systems requires a supervision system that includes coordination as well as fault detection and diagnosis capabilities. The objective of coordination system is to enforce the plant to legal and safe behavior, while the objective of the fault detection and isolation is to detect and isolate the failures according to the observation traces generated by the system. This paper proposes a new supervisory control technique for online fault detection, isolation and coordination based on Petri net models. This technique has two components, a supervisor, and a diagnoser. The latter evaluates the faulty states without checking the properties that restrict the performance of the conventional dignosers to certain types of Petri net models. This paper also presents two-robot surveillance system, as an application example, to test the proposed algorithm.

Research Situation Analysis of On-line Detection Technology for Power Transformer Winding Deformation

Background: Winding deformation is one of the most common faults in power transformers, which seriously threatens the safe operation of transformers. In order to discover the hidden trouble of transformer in time, it is of great significance to actively carry out the research of transformer winding deformation detection technology. Methods: In this paper, several methods of winding deformation detection with on-line detection prospects are summarized. The principles and characteristics of each method are analyzed, and the advantages and disadvantages of each method as well as the future research directions are expounded. Finally, aiming at the existing problems, the development direction of detection method for winding deformation in the future is prospected. Results: The on-line frequency response analysis method is still immature, and the vibration detection method is still in the theoretical research stage. Conclusion: The ΔV − I1 locus method provides a new direction for on-line detection of transformer winding deformation faults, which has certain application prospects and practical engineering value.