Automated Fault Detection Methods Research Articles

Fault detection methods for 2D and 3D geomechanical numerical models

AbstractWe develop automated methods for fault detection utilizing static stress and deformation fields at the onset of failure derived from numerical analysis. We calculate combinations and normalization of the distance from the Mohr circle to the Coulomb envelope, and of the second deviatoric strain invariant. A variation of the Cauchy distribution of these fields allows us to focus on the low values indicating rupture, with the help of the scale parameter δ. A threshold is then applied to decide at each spatial node of the mesh whether the material has reached failure or not. We then determine fault lines and planes from these isolated failure zones using image processing techniques, such as the Hough and the Radon transforms, or through a combined approach involving automated sorting of the nodes reaching failure through the k‐means clustering technique followed by polynomial fitting to retrieve analytic expressions of the fault curves (in 2D) or fault surfaces (in 3D). The methods are efficient except when the stress field results in diffuse rupture zones that do not localize onto fault surfaces despite tuning δ. We also highlight the advantages of using the combined clustering/poly‐fitting approach for 3D models compared to the image processing techniques. These automated fault detection methods should be useful in the interpretation of diverse failure mechanisms obtained through parametric sensitivity analyses requiring hundreds of simulations. The stress and strain fields used were derived from a numerical implementation of limit analysis, but classical finite‐difference or finite‐element techniques could have been used.

Fault detection method based on an automated operating envelope during transient states for the large turbomachinery

In the energy generation business steam powered turbo-generators still play an important role in electrical power generation all over the world. Every facility using steam turbines considers them as the critical machinery. Such machines should be well-maintained, properly handled, and precisely diagnosed in order to achieve the best performance and safety. The most valued data about the technical health are collected during machine’s shut-downs and run-ups. These data are more than seldom and hard to assess without expert’s knowledge with strong theoretical background and experience. Main novelty of the paper is the automated method for novelty detection of machine’s vibration. Most proposed methods apply to smaller machines with rolling bearings, whereas we propose the method for large machines with sliding bearings, which have much different behavior. The application of the method is support of the plant maintenance staff to evaluate deviations of turbo-sets from a healthy state based on the concept which we called the Operating Envelope. The envelope is created based on the data from a vibration sensor during the transient state. In this paper we consider a single vibration sensor and only the first harmonic amplitude of this signal. To set the acceptance limits within which turbo-set’s dynamic response will be considered as acceptable, we used the cubic spline interpolation coupled with expert judgement. Beyond these limits the state of the turbo-set is considered as unhealthy, so it is an automated fault detection method. In such a case a machine should be a subject to further and deeper diagnostic analysis. The method was validated on the data from the 13K242 type (a 200 MW class turbine) steam turbine. We also proposed a set of parameters to evaluate the severity of malfunction.

Proposal of noninvasive failure diagnosis of electrical motor using googlenet

Fault diagnosis is a useful tool that reduces system maintenance risks and costs. However, data related to the system's nominal and fault operating behavior is often not collected and stored adequately, it is difficult to identify and suggest automated fault detection methods. This study proposes a solution to apply deep learning technique on the convolutional neural network (CNN) to identify some common errors on induction motors based on operation sound. The opreration sound signal emitted from on a 0.37 kW two-pole induction motor is collected in some cases such as normal operation, phase loss, phase difference and bearing breakage. Their 2-D scalogram images are analyzed by continuous Wavelet transformation which is used to train and evaluate the deep learning CNN (i.e. GoogLeNet) to identify the above faults. Experimental results show that this method can diagnose induction motor faults with accuracy up to 98.8%.

Detecting loss-of-coolant accidents without accident-specific data

This paper develops an automated fault detection tool to detect very small LOCAs in pressurized water reactors that would be difficult for operators to detect manually. One of the primary challenges with previous automated fault detection methods, which are data-driven, is that they require data from LOCAs; however, it may be difficult to capture real operational data from LOCA scenarios. This work uses a physics-inspired approach that equates the physical effects of a LOCA to changes in known variables. This approach enables the detection of very small LOCAs using data-driven approaches that use nominal operating data without the need for LOCA data. The approach combines data-driven modeling with control-theoretic estimation techniques to detect LOCAs and estimate their magnitudes in real-time. First, simulated process data for a variety of nominal operating conditions is collected using a generic pressurized water reactor simulator. Then, that data is used to train an artificial neural network regression model that captures the nonlinear plant dynamics. Finally, the regression model is used in a particle filter to detect the onset and estimate the magnitude of the leak. These methods are successfully verified using LOCA simulations that would be hard to manually distinguish from normal operating transients.

Novel real-time model-based fault detection method for automatic identification of abnormal energy performance in building ventilation units

Studies show that buildings are able to attain energy savings of up to 30% by rectifying faulty HVAC systems using fault detection and diagnostic methods. In this study, we propose an automated top-down model-based fault detection method that can detect abnormal electricity consumption of building ventilation systems. This is achieved by employing a dynamic building model of the highly efficient OU44 teaching building at the University of Southern Denmark in Odense, allowing for identification and quantification of sub-optimal performance in real-time. The methodology is based on a statistical definition of abnormal telecommunicational operation that employs an approximation of the Chernoff bound proposed by Cheung et al. We develop and implement the Chernoff bound method for building ventilation systems and propose a simple approach of defining threshold limits based on a percentage of the model predicted performance of the ventilation systems. Each time the observed data deviates from the threshold limits, a suspicion is then evoked. The urgency of each suspicion period is classified into three levels of urgency (high, medium and low) depending on the probability of abnormality calculated from the Chernoff approximation, thereby aiding building managers in resource delegation matters. A FDD analysis for a time period of two months is conducted using the proposed method. The results show some high urgency classified suspicion periods within the investigated period, which are manually diagnosed by utilising room level sensor data. This led to the observation of a faulty occupancy counter. The proposed method shows promising outlook within the field of automated fault detection methods in terms of highlighting periods with abnormal behaviour.

Automated valve fault detection based on acoustic emission parameters and support vector machine

Reciprocating compressors are one of the most used types of compressors with wide applications in industry. The most common failure in reciprocating compressors is always related to the valves. Therefore, a reliable condition monitoring method is required to avoid the unplanned shutdown in this category of machines. Acoustic emission (AE) technique is one of the effective recent methods in the field of valve condition monitoring. However, a major challenge is related to the analysis of AE signal which perhaps only depends on the experience and knowledge of technicians. This paper proposes automated fault detection method using support vector machine (SVM) and AE parameters in an attempt to reduce human intervention in the process. Experiments were conducted on a single stage reciprocating air compressor by combining healthy and faulty valve conditions to acquire the AE signals. Valve functioning was identified through AE waveform analysis. SVM faults detection model was subsequently devised and validated based on training and testing samples respectively. The results demonstrated automatic valve fault detection model with accuracy exceeding 98%. It is believed that valve faults can be detected efficiently without human intervention by employing the proposed model for a single stage reciprocating compressor.

Review of long-term fault detection approaches in solar thermal systems

Abstract This paper presents an overview, assessment and comparison of automated fault detection methods that check if solar heating systems are functioning correctly. Fault detection in solar thermal systems is important to minimize the time when the system is not functioning well, thereby ensuring an optimal energy (and economic) yield. During the past decades many systems have been monitored, mainly for scientific or demonstration projects by logging measurement data which was subsequently analysed by an expert. Automation of fault detection is necessary to reduce costs and minimize experts’ time needed for analysis of a system. An overview of existing fault detection approaches is given; these are evaluated and compared with a multi-criteria analysis. The only commercially available automated method, the Input–Output Controller, detects faults causing more than 20% energy loss in the solar loop. The function control approach is cheap without a heat meter, and only relies on few sensors to check how several components in the solar loop are functioning with algorithms. The approach developed at Kassel University checks how well a solar plant is functioning both with plausibility checks and with energy balances based on simulations. This method includes a larger part of the solar heating system and therefore requires more measurement equipment. Further research and application of several fault detection methods should improve the effectiveness and costs of these methods.