Efficient Fault Detection Research Articles

Performance-Driven Distributed PCA Process Monitoring Based on Fault-Relevant Variable Selection and Bayesian Inference

Multivariate statistical process monitoring involves dimension reduction and latent feature extraction in large-scale processes and typically incorporates all measured variables. However, involving variables without beneficial information may degrade monitoring performance. This study analyzes the effect of variable selection on principal component analysis (PCA) monitoring performance. Then, it proposes a fault-relevant variable selection and Bayesian inference-based distributed method for efficient fault detection and isolation. First, the optimal subset of variables is identified for each fault using an optimization algorithm. Second, a sub-PCA model is established in each subset. Finally, the monitoring results of all of the subsets are combined through Bayesian inference. The proposed method reduces redundancy and complexity, explores numerous local behaviors, and provides accurate description of faults, thus improving monitoring performance significantly. Case studies on a numerical example, the Tennessee Eastman benchmark process, and an industrial-scale plant demonstrate the efficiency.

COMPUTER VISION SYSTEM FOR AUTOMATIC CERAMIC TILES INSPECTION

Ponte Academic JournalSep 2016, Volume 72, Issue 9 COMPUTER VISION SYSTEM FOR AUTOMATIC CERAMIC TILES INSPECTIONAuthor(s): POLEPOGU RAJESH ,KUNDURU UMAMAHESWARIJ. Ponte - Sep 2016 - Volume 72 - Issue 9 doi: 10.21506/j.ponte.2016.9.16 Abstract:Quality control is an important issue in the ceramic tile industry. On the other hand maintaining the rate of production with respect to time is also a major issue in ceramic tile manufacturing. Due to massive production of ceramic tiles, using human resources for defect detection is quite tedious and time consuming process. So we need to automate the process. Various optical technology based approaches came into play such as Automatic Optical Imaging. It is effective but very expensive technique for small-scale manufacturers. Therefore, many researchers started developing image processing algorithms for defect detection in ceramic tiles using less expensive cameras. In this paper we proposed an automated computer vision system with efficient defect detection technique, which can detect surface defects like blobs and cracks on tiles with high accuracy and fast response. This automated method helps us to acquire knowledge about the pattern of defect within a very short span of time. To automate the whole process, we have developed a ‘Graphical User Interface’ (GUI) in MATLAB using GUIDE (Graphical user interface design environment). We have used Sobel operator for edge detection and image subtraction techniques for fault detection in ceramic tiles. Index Terms— CV (Computer Vision), GUI (Graphical User Interface), GUIDE (Graphical user interface design environment), Threshold segmentation. Download full text:Check if you have access through your login credentials or your institution Username Password

Frequency Dependence of the Defect Sensitivity of Guided Wave Testing for Efficient Defect Detection at Pipe Elbows

Frequency Dependence of the Defect Sensitivity of Guided Wave Testing for Efficient Defect Detection at Pipe Elbows

Defect Detection in Tire X-Ray Images Using Weighted Texture Dissimilarity

Automatic defect detection is an important and challenging problem in industrial quality inspection. This paper proposes an efficient defect detection method for tire quality assurance, which takes advantage of the feature similarity of tire images to capture the anomalies. The proposed detection algorithm mainly consists of three steps. Firstly, the local kernel regression descriptor is exploited to derive a set of feature vectors of an inspected tire image. These feature vectors are used to evaluate the feature dissimilarity of pixels. Next, the texture distortion degree of each pixel is estimated by weighted averaging of the dissimilarity between one pixel and its neighbors, which results in an anomaly map of the inspected image. Finally, the defects are located by segmenting this anomaly map with a simple thresholding process. Different from some existing detection algorithms that fail to work for tire tread images, the proposed detection algorithm works well not only for sidewall images but also for tread images. Experimental results demonstrate that the proposed algorithm can accurately locate the defects of tire images and outperforms the traditional defect detection algorithms in terms of various quantitative metrics.

ETS (Efficient, Transparent, and Secured) Self-healing Service for Pervasive Computing Applications

To ensure smooth functioning of numerous handheld devices anywhere anytime, the importance of self-healing mechanism cannot be overlooked. Incorporation of efficient fault detection and recovery in device itself is the quest for long but there is no existing self-healing scheme for devices running in pervasive computing environments that can be claimed as the ultimate solution. Moreover, the highest degree of transparency, security and privacy attainability should also be maintained. ETS Self-healing service, an integral part of our developing middleware named MARKS (Middleware Adaptability for Resource discovery, Knowledge usability, and Self-healing), holds promise for offering all of those functionalities.

Diagnostic Bayesian networks for diagnosing air handling units faults – Part II: Faults in coils and sensors

This is the second part of a study on diagnostic Bayesian networks (DBNs)-based method for diagnosing faults in air handling units (AHUs) in buildings. In this part, 4 DBNs are developed to diagnose faults in heating/cooling coils, sensors and faults in secondary supply chilled water/heating water systems. There are 18 typical faults concerned and 35 fault detectors introduced. The DBNs are developed mainly on the basis of first principles and fault patterns resulted from literature and three AHU fault detection and diagnosis (FDD) projects. Efficient fault detection rules/methods from a comprehensive literature survey are integrated into the DBNs. Also, some new fault detection rules are developed. The 4 DBNs were evaluated using experimental data from ASHRAE Project RP-1312. Results show that the proposed DBNs effectively diagnosed AHU faults.

Integration of multivariate statistical process control and engineering process control: a novel framework

Statistical process control is being used along with classical feedback control systems (which are also termed as Engineering Process Control, EPC) for the purposes of detecting faults and avoiding over adjustment of the processes. This paper evaluates the effectiveness of integrating SPC with EPC for both fault detection and control. A novel framework for fault detection using Multivariate Statistical Process Control (MSPC) is proposed here and illustrated with a case study. The simultaneous application of MSPC control charts to process inputs and outputs or in other words “joint monitoring” of process inputs and outputs is shown here to provide efficient fault detection capabilities. An example of Heating Ventilation and Air Conditioning (HVAC) systems is simulated here and used as a case study to demonstrate the detection capabilities of the proposed framework. Moreover, the capabilities of the proposed framework were enhanced by inclusion of a corrective action scheme, thus leading to a complete control system with fault detection and correction.

A deeper insight into fault location on long submarine power cables

This paper provides a deep insight into fault location on long Submarine Power Cables. Several field results on submarine power cable faults are provided. For land cables the main focus is on the electrical and thermal design of cable insulation and on the electro-thermal and thermo-mechanical design of cable accessories, in order to grant sufficient endurance performances and reliability to the whole cable systems (Mazzanti, G., Marzinotto, M., Extruded cables for high-voltage direct current transmission: advances in research and development. New York, Wiley-Blackwell, 2013). In submarine cables more issues arise, some related to the harsh laying environment, some others—even more troublesome—associated with man-made activities. Indeed, on the one hand submarine power cables are subject to strong mechanical stresses during the laying operations and critical service conditions in their working ambience. On the other hand, submarine cables are continuously exposed in all water depth to random mechanical damages caused by fishing gears, anchors and natural hazards (Bawart, M., Marzinotto, M., in Insulated Conductors Committee ICC/IEEE-PES, C11D submarine cables, St. Petersburg, FL, November, 2012; CIGRE, Third-party damage to underground and submarine cables, Brochure 398, WG B1.21, December, 2009). The longer the expected life and the longer the path of the submarine cable link, the higher is the probability of facing one or more faults due to human activities. Based on surveys about submarine cable failure data recorded worldwide over long periods, it can be concluded that the probability of experiencing at least one fault during lifetime is close to certainty for long submarine links. Statistically most damages to submarine cables are caused by human activities; only a low percentage is caused by natural hazards. Based on growing energy demand and dependency on offshore produced renewable energy, submarine power cables become essential for reliable electric power supply and often can be classified as critical infrastructure (International Cable Protection Committee—ICPC, About submarine power cables, 2011, www.iscpc.org ). Repair of damaged submarine power cables requires specialized ships as well as experts to recover the cable from the sea bed and replace the faulty cable section. Another critical aspect associated with long submarine cables is that, whenever a fault occurs, a fairly long time is spent for repair. For this reason, fast and efficient fault detection is essential in order to reduce the overall outage time as much as possible. All these aspects are discussed in this paper. The best practice commonly employed for classifying submarine power cable fault types are included in the paper, together with the results of measurements carried out in the field. The paper points out that fault location on submarine power cables differs by much from classical cable fault location on buried land cables as to both conditions and measuring methods, thereby illustrating the most efficient cable fault location methods. Some field results on submarine power cable faults are provided, measured on AC submarine cables as well as on HVDC submarine links. A unique case study of fault location on longest HVDC Submarine Link will illustrate TDR based measurements on cable lengths above 400 km. The case studies further focus on TDR diagram analysis in order to explain how to identify cable joints. The results prove that the overall outage time for repair activities can drop significantly if the fault location system is peculiarly designed for detecting faults in very long submarine cables with a good measuring accuracy. The hazards for operators and instruments connected to the huge amount of electrical energy that may be stored in very long links are also tackled in the paper, thereby addressing the particular safety issues involved by extra-long submarine cables.

Quality-Related Fault Detection in Industrial Multimode Dynamic Processes

Multivariate statistical process monitoring (MSPM) methods are powerful tools for detecting faults in industrial systems. However, industrial processes are often subjected to dynamic changes. This dynamic behavior is mainly due to set-point changes and nonlinearities. Because of the nonlinearity of processes, the performance of the classical MSPM methods, which are mainly based on the linearity assumption, becomes unsatisfactory, since the process characteristics will change from one operating point to another. The main objective of the work is to develop an efficient fault detection technique for complex industrial systems, using process historical data and considering the nonlinear behavior of the process. In the proposed approach, the nonlinear system is assumed to be linear around the operating points and therefore considered as a piecewise linear system corresponding to each operating mode. The performance and effectiveness of this approach are demonstrated using data obtained from a paper machine and compared with an available method.

Efficient fault detection using support vector machine based hybrid expert system

This paper demonstrates the methodology of fault classification of rotating machinery using support vector machine (SVM) in combination with genetic algorithm and particle swarm optimization. In order to detect the machine health condition, classifier uses the features as the inputs from the preprocessed raw signal of a machine. Support vector machine classifier prepared in combination of hybrid adaptive particle swarm optimization and adaptive genetic algorithm (HAPAG) proposed for proficient flaw detection. An industrial case study of a centrifugal pump is considered and the data is given for both training and testing of the classifier. A similar study with comparable existing fault classifiers on the identification triumph is investigated. SVM based HAPAG system results in clustering the various faults with more than 90 % accuracy when compared with adaptive tuning of SVM based techniques like SVM—adaptive particle swarm optimization and SVM—adaptive genetic algorithm. The outcome indicates the adequacy of choosing the classifiers in finding the machine health condition.

A Variability-Aware Adaptive Test Flow for Test Quality Improvement

In this paper, we propose a process-variability-aware adaptive test flow that realizes efficient and comprehensive detection of parametric faults. A parametric fault is essentially a malfunction in a large-scale integration chip, which is caused by the variability in fabrication processes. In our adaptive test framework, test pattern sets are altered on individual chips in order to apply the optimal set of test patterns for each chip, and thus the test coverage is improved and the test time is reduced. The test pattern is chosen on the basis of parameter estimations measured using an on-chip sensor with respect to statistical timing information. We also propose a novel metric to quantize the test coverage suitable for evaluating the test quality of parametric faults. Our experimental results using an industrial design show that the proposed flow significantly improves the parametric fault coverage and test efficiency compared to conventional test flows.

Built in Self-Test for 4 × 4 Signed and Unsigned Multipliers in FPGA

ASTRACT Built in self-test (BIST) is a technique or a method which allow the circuit to test itself. BIST increases the controllability and observability of integrated circuit therefore it is easier to apply inputs and then detect faults from it [11]. BIST also decreases the time of testing integrated circuits & gives very high fault coverage. Therefore in many ways BIST help us in detecting fault in integrated circuits. This paper presents an efficient fault detection algorithm for 4 × 4 signed & unsigned multiplier in field programmable gate array (FPGA). These techniques were successfully applied on booth, braun & unsigned array multipliers.

On Selection of Modulus of Quadratic Codes for the Protection of Cryptographic Operations against Fault Attacks

Quadratic residue codes are introduced as an effective and efficient fault detection technique to protect cryptographic devices against fault attacks. In this paper, we re-consider these codes in an adversarial model, where a powerful attacker can introduce faults with high precision and accuracy. We present two analysis techniques that can lead to successful attacks against quadratic codes if the modulus is not chosen carefully. We provide in-depth theoretical analysis that covers wide range of attacks and present results of practical concerns such as exact number of undetected faults and effective countermeasures. Our analysis is generic in the sense that it can be extended to other residue codes.

Data set preprocessing methods for the artificial intelligence-based diagnostic module

The paper presents the application of statistical (econometrics-originated) methods to process learning and testing data sets used by the artificial intelligence (AI) methods in the diagnostics of analog systems. Before the training and evaluation of the intelligent module is performed, the measurement data are analysed to minimize the number of attributes (symptoms) required to distinguish between different states of the System Under Test (SUT). This way the knowledge extracted from the set is simplified, increasing the operation speed and minimizing the threat of overlearning. Also, elimination of unnecessary symptoms from the set allows for decreasing the set of test points where measurements are taken (which is economically desirable). Preprocessing operations include elimination of constant or quasi-stationary symptoms and finding their minimal set, allowing for the efficient fault detection or parameter identification. The paper focuses on the Hellwig and Multiple Correlation Coefficient methods adjusted to the technical diagnostics applications. They are implemented to optimize data sets obtained from simulation of the fifth order lowpass filter. Their usefulness is tested using the artificial neural network (ANN) and Rough Sets (RS) classifiers responsible for detection, and identification of parametric faults.

Fault Management Using Cluster-Based Protocol in Wireless Sensor Networks

Fault tolerance has been identified as key challenges in the design and operations of Wireless Sensor Network (WSNs). Failures are inevitable in wireless sensor networks due to inhospitable environment and unattended deployment. Therefore, it is necessary to detect the networks for recovery from the failure to sustain it. WSNs are self-organized using clustering algorithms to conserve energy. LEACH (Low Energy Adaptive Clustering Hierarchy) protocol is one of the significant protocols for routing in WSN. In LEACH, sensor nodes are organized in several small clusters where there are cluster heads in each cluster. These CHs gather data from their local clusters aggregate them & send them to the base station. The proposed scheme is supposed to be an efficient fault detection and recovery mechanism to make the network fault-tolerant and achieve reliability and quality of service. Performance analysis and evaluation will be made using several scenarios of wireless sensor networks.

LATEST PROGRESS OF FAULT DETECTION AND LOCALIZATION IN COMPLEX ELECTRICAL ENGINEERING

Abstract In the researches of complex electrical engineering, efficient fault detection and localization schemes are essential to quickly detect and locate faults so that appropriate and timely corrective mitigating and maintenance actions can be taken. In this paper, under the current measurement precision of PMU, we will put forward a new type of fault detection and localization technology based on fault factor feature extraction. Lots of simulating experiments indicate that, although there are disturbances of white Gaussian stochastic noise, based on fault factor feature extraction principal, the fault detection and localization results are still accurate and reliable, which also identifies that the fault detection and localization technology has strong anti-interference ability and great redundancy.

An Efficient Multiple Fault Detection and Isolation Scheme for Discrete Event System

The malfunction of sensors, actuators, and erroneous actions of human operators can have some disastrous consequences in high risk systems especially if these systems have multiple faults that can lead to undesirable shutdowns and consequently a plant output reduction. This paper introduced a new scheme to detect and isolate multiple faults in complex discrete event systems. This scheme comprises a set of discrete rules and an inference engine. The former characterizes the dynamic behavior of a process that is modeled by a Petri net and the latter infers the consequent part of the fired rules. The consequent parts are aggregated and compared with the original Petri net model of the process for detecting multiple faults. The proposed inference engine follows the forward reasoning strategy that fires transitions from the initial marking until the final marking is reached. Industrial processes are employed for testing the soundness of the proposed scheme.

Commissioning and Operational Control of Photovoltaic Power Plants through Online Simulation

Photovoltaic system simulation tools are mostly used for the planning and yield prediction of power plants.During the commissioning procedure and especially during the operation of a photovoltaic power plant simple algorithms are used to convert the measured power under varying irradiance conditions to standard test conditions. Faults are only detected, if there is a significant deviation between the photovoltaic system power or daily production and the estimated production based on measured irradiance levels. In large PV power plants this can result in high economic losses, which could be reduced if the monitoring mechanisms were supported by precise modelling of the photovoltaic generator.In this work, an integrated monitoring and simulation method was developed and tested for very precise online simulation of photovoltaic power plants. The method is based on detailed one or two diode current-voltage curve modelling of the photovoltaic generator. Depending on the technology, between six (crystalline) and ten (thin films) modelling parameters for the IV curve are required. The parameters are either derived from the data sheet information of the PV modules or more precisely from measured IV curves of the modules at difference irradiance levels.Using simulated IV curves, an MPP tracking algorithm and voltage dependent inverter models, a simulation accuracy of 2% is reached, so that any problems with the PV generator can reliably be detected. Using internet based communication technology for the data exchange, the online simulation can be carried out directly on the site of the photovoltaic power plant or from any remote location. Remote simulation offers the advantage to the PV plant operator, that it can also be done as a simulation service and that the software environment does not need to be purchased.To demonstrate the method, online simulation of several grid connected PV generators up to the MW power range is carried out during the commissioning and operation of photovoltaic plants. A powerful visualisation within the same software environment includes both the monitored and the simulated online data sets, so that a very efficient fault detection scheme is available. The method is excellently suited to provide automatic and real time fault detection and significantly improve the commissioning procedure for photovoltaic plants of all sizes.

Improving the Diagnostic Performance for Dynamic Systems through the use of Conflict-Driven Model Decomposition

This work studies potential ways of integration of two techniques for fault detection, isolation, and identification in dynamic systems: Lydia-NG suite of diagnosis algorithms and Consistency-based Diagnosis with Possible Conflicts. By integrating both techniques, Lydia- NG will benefit from a more efficient fault detection and isolation task, and Possible Conflicts will benefit from the identification capabilities of Lydia-NG. In this paper, we define a common framework that integrates both techniques, and then we apply the proposed integrated approach to a three-tank system, and draw some conclusions about potential ways of integration.

Electronic Returnless Fuel System Fault Diagnosis and Isolation: A Data-Driven Approach


 
 
 The Electronic Return-less Fuel System (ERFS) manages the delivery of fuel from the fuel tank to the engine. The pressure in the fuel line is electronically controlled by the fuel system control module by speeding up or slowing down the fuel pump. This allows the system to efficiently control the amount of fuel provided to the engine when compared to vehicles equipped with a standard fuel system wherein the fuel pump continuously runs at full speed. A failure in the fuel system that impacts the ability to deliver fuel to the engine will have an immediate effect on system performance. Consequently, improved reliability and availability, and reduction in the number of walk-home situations require efficient fault detection, isolation and prognosis of the ERFS system. This paper develops and implements data-driven fault detection, isolation and severity estimation algorithms for the ERFS. The HIL Fuel System Rig and a Chevrolet Silverado truck were used to collect and analyze the fuel system behavior under different fault conditions. Several data-driven classifiers, such as support vector machines, K- nearest Neighbor, Discriminant analysis, Bayes classifier, Partial- least squares, Quadratic and Linear classifiers, were implemented on a limited set of data for both training and testing. Regression techniques, such as Partial least squares regression and Principle component regression, are used to estimate the severity of faults. The resulting solution approach has the potential to be applicable to a wide variety of systems, ranging from automobiles to aerospace systems.