Smart Fault Detection Research Articles

An Intelligent Fault Detection Method of Industrial Gearboxes With Robustness One-Class Support Matrix Machine Toward Multisource Nonideal Data

Various kinds of sensing data can be acquired for smart fault detection. Each signal source has spatial attributes and strong correlations exist between different data sources. However, most of the existing fault detection models are established in the vector domain, which would destroy the structure information embedded within multisource data. Besides, the nonideal data, especially strong-noise data and mislabeled data, will seriously affect the fault detection performance. Therefore, a matrix-form one-class model called robustness one-class support matrix machine (ROCSMM) is proposed for industrial gearbox fault detection under multisource nonideal data. ROCSMM extends the traditional one-class support vector machine to the matrix domain, which can retain the topological structure information of multisource data while improving the fault detection performance. Besides, an adaptive weight generation strategy is designed for ROCSMM according to the prior distribution of matrix samples. This strategy can eliminate the negative impact of nonideal data on ROCSMM to the greatest extent and improve the model's robustness. The experimental results indicate that the proposed model is superior to other cutting-edge models in the fault detection of industrial gearboxes.

Computational sensor nodes optimization for smart anomaly detection applied to wind energy

The increasing use of wind energy generators in mountainous areas is an important necessity in establishing and enhancing renewable energy. This procedure has created a need for reliable and efficient methods for detecting anomalies remotely basing on computational sensor nodes optimization.These methods are useful for detecting deviations in the generator's behavior that may indicate a potential failure, allowing for prompt maintenance, minimizing the intervention costs, and ultimately reducing downtime and improving the generator's overall reliability compared to previous techniques. In this context, this paper presents the principal smart anomaly detection methods applied to wind energy generators. It covers different techniques, including machine learning algorithms, statistical approaches, and deep learning models. Furthermore, this paper aims to perform the optimization of computational sensor nodes (CSN) as new strategy, instead of basic wireless sensor network, in smart fault detection. This paradigm is growing very rapidly in varied fields. One of the promising application fields of CSN is wind energy generators, where this technology offers important support. This technology will allow for precise resource management. The experimental results show that the proposed approach can effectively detect anomalies in wind turbine data, and it outperforms traditional statistical methods with 99 % recognition rate against 93 % in the best way..

Smart fault detection of HTS coils using artificial intelligence techniques for large-scale superconducting electric transport applications

Superconducting coil is an essential and critical component in any superconducting apparatus used in large-scale power applications such as in superconducting machines of propulsion systems, in fault current limiters of the distribution system for future cryo-electric aircraft, or in windings of superconducting transformers for power grid applications. The superconducting coils in winding of large-scale power devices operate in kind of harsh environments from both temperature—considering liquid hydrogen or gaseous helium as coolant—(thermal stress) and electro-magneto-mechanical stress, point of views. Reliable operation of the coils in winding is of vital importance for the reliability of the superconducting device and the safety of the application that the device is used for. If the superconducting coil confronts a fault or an abnormal condition in the laboratory-level operation, it is straightforward to test the coil by measuring its critical current, AC loss, etc, to find whether it is damaged or not. However, there would be an urgent need to have faster and more intelligent fault detection and condition monitoring approaches with the possibility to become fully autonomous and real-time, in large-scale power applications, especially in sensitive applications such as in future cryo-electric aircraft, or in the fusion industry. To reach such intelligent fault-finding approaches, artificial intelligence-based techniques have been foreseen to be a promising solution. In this paper, we have developed an intelligent fault detection technique for finding a faulty superconducting coil, named the frequency-temporal classification method. This method has two main steps: first, this paper utilizes the discrete Fourier transform and independent component analysis to convert measurement signals of the healthy and faulty coils from (1) the time-series domain to the frequency domain; and (2) into time-series source signals. Second, this paper trains the support-vector machine using the derived frequency components. This trained model is then used for making fault detection for other superconducting coils. The developed technique can classify a fault with 99.2% accuracy.

Deep Neural Network with Hilbert–Huang Transform for Smart Fault Detection in Microgrid

The fault detection method (FDM) plays a crucial role in controlling and operating microgrids (MGs), because it allows for systems to rapidly isolate and restore faults. Due to the fact that MGs use inverter-interfaced distributed production, conventional FDMs are no longer appropriate because they are dependent on substantial fault currents. This study presents a smart FDM for MGs based on the Hilbert–Huang transform (HHT) and deep neural networks (DNNs). The suggested layout aims to prepare the fast detection of fault kind, phase, and place data to protect MGs and restore services. The HHT pre-processes the branch current measurements obtained from the protective relays to extract the characteristics, and singular value decomposition (SVD) is used to extract some features from intrinsic mode functions (IMFs) that are obtained from HHT to use as input of DNNs. As part of the fault data development, all the information eventually enters the DNNs. Compared with prior studies, this suggested method provides considerably superior fault-type identification accuracy. It is also possible to determine new fault locations. A detailed assessment analysis of this suggested FDM was conducted on IEEE 34-bus and MG systems to demonstrate its effectiveness. The simulations indicated that the proposed method is effective for detecting precision, computing time, and robustness to measurement uncertainties.

Triboelectric pad journal bearing for self-powered condition monitoring

With the widespread application of the Internet of Things (IoT) technology, smart fault detection and condition monitoring are urgently required in mechanical systems. The emergence of triboelectric nanogenerators (TENG) provides a novel strategy to achieve self-powered sensing, which can serve as an ideal candidate for the condition monitoring of bearings. In this study, a triboelectric pad journal bearing (T-PJB) was designed and fabricated to directly monitor the bearing condition. Through simulation, the geometrical structure of the T-PJB for high output was optimized, and the electrical performance was examined under various rotating conditions. The speed of the bearing shaft could be accurately measured using time- and frequency-domain methods. During the operation, faults caused by unbalanced forces were effectively identified based on the characteristics of the electrical output signals. Furthermore, the proposed T-PJB can be applied to monitor lubrication states because of its sensitive response to different lubrication conditions. Condition monitoring using the proposed T-PJB greatly contributes to reduced maintenance costs and long-term stable operation of various sliding bearings. Furthermore, the demonstrated smart T-PJB has promising feasibility in industrial applications for rotating machinery, establishing the foundation for smart plants.

An Efficient LoRa-Enabled Smart Fault Detection and Monitoring Platform for the Power Distribution System Using Self-Powered IoT Devices

Transient stability and supply disturbances are common yet unwelcome phenomena in power distribution systems, particularly in sub-Saharan Africa. The growing demand for greater reliability and dependability in power delivery has aroused the interest of researchers and renewed the pursuit of advanced technological solutions for fault detection and location determination at medium and low-voltage levels. The length of the distribution network typically ranges from hundreds to thousands of kilometers. In this regard, the management of distribution networks, including the identification of faulty segments, is a significant recurrent challenge facing power-system operators. With the ever-expanding distribution network and regulatory demands for service reliability, the challenge for network operators is daunting. However, the deployment of IoT technologies in the energy distribution infrastructure would significantly accelerate the detection and location of faults, thus transforming the electricity delivery service into one that is responsive, communicative, attractive, and robust. This study proposes, designs, and implements a reasonably priced LoRaWAN-based IoT platform for monitoring distribution networks. The study was conducted in Nakuru County, Kenya on an actual and active distribution network owned and managed by Kenya Power Company. Experimental results showed that a trigger is generated at the network-monitoring center in about 100 msec of the occurrence of a fault on the distribution network, thus enabling quick, prompt, and immediate commencement of reparative action. Furthermore, practical evaluation has shown that this platform is well adapted for the context of developing countries where budgetary constraints and cost prohibitions hinder the upgrade of the legacy grid into fully-fledged smart entities.

Distributed optical fibre sensor for condition monitoring of mining conveyor using wavelet transform and artificial neural network

Condition monitoring of mining conveyors is highly essential, not only to reduce economic loss due to sudden component breakdown but also to prevent safety risks to maintenance personnel. The vibration measurement by means of Distributed Optical Fibre Sensors (DOFS) is a potential solution to provide cost-effective real-time monitoring of mining conveyors. However, an effective application of detecting early damage and damage progression mainly depends on collection quality of optical signal and robust signal processing methods. Therefore, this study focuses on the identification of an effective signal processing method for real-time monitoring of mining conveyor systems. A prototype conveyor system with various levels of idler damages was built to run experimental tests and to collect optical signals for further analysis. A wavelet-based damage detection method is proposed, and its efficiency is evaluated. Additionally, Artificial Neural Network (ANN) is integrated to develop a smart fault detection technology for fast detection and damage classification. The proposed method has 99% accuracy in classify damage condition under varying coveyor operating speed.

Research on the application of Decision Tree and Random Forest Algorithm in the main transformer fault evaluation

The detection and monitoring methods of power equipment are relatively complex. The analysis of a large amount of data obtained from detection and monitoring and the development of data recording methods make it possible to obtain higher-dimensional power data. If we adopt appropriate data analysis method to the data sets, we can better get the potential laws and value of power data. In this paper, we took various basic monitoring data and faults records of the 110kV main transformers into consideration, and the grid-equipment-environment data was fused. Based on the fused data, we used the decision tree and random forest algorithm to evaluate the main transformers’ defects and faults. The evaluation results of the two algorithms were obtained and compared, which proved the effectiveness of the fault evaluation algorithm and selected a more accurate fault evaluation algorithm. This paper provides new ideas for smart fault detection for power grid, and provides a reference for a more in-depth evaluation of power grid equipment.

A smart fault-detection approach with feature production and extraction processes

In this paper, a smart fault-detection approach with feature production and extraction procedures is developed for industrial ball bearing systems. By developing a set of chaotic-mapping systems composed of a main system and data-feeding system with appropriate parameters, the vibration signals of different fault states captured from ball bearings in the time domain can be mathematically mapped to the chaotic domain for feature production. Furthermore, through the designed feature extraction process, the relevant Euclidean feature values (EFVs) can be obtained for the classification of four different fault states. Three fault conditions at diameters of 7 mil, 14 mil, and 21 mil and a depth of 0.011 inches are illustrated for performance investigations. The experimental results show that the proposed smart detection approach is effective and feasible for identifying different fault states in real time.

Smart Fault-Detection Machine for Ball-Bearing System with Chaotic Mapping Strategy.

In this paper, a set of smart fault-detection approach with chaotic mapping strategy is developed for an industrial ball-bearing system. There are four main statuses in this ball-bearing system: normal, inner race fault, outer race fault, and ball fault. However, it is hard to simply classify each of them through their vibration signals in time-series. By developing a nonlinear error dynamic system as well as a chaotic mapping strategy, the signals in the time series can be converted into the chaotic domain, which are revealed in 3D phase portraits. Further, through collocation of clustering methods, such as Euclidean distance (ED) and the kernel method of K-means (KM), the proposed 3D phase portraits of each different state can be efficiently identified through checking the autonomously adjusted ranges of feature values. The experiment results show that the proposed smart detection approach is effective and feasible, and the accuracy of detection in the testing stage is close to 100%.

Wavelet-Multi Resolution Analysis Based ANN Architecture for Fault Detection and Localization in DC Microgrids

DC microgrids present an effective power system solution for increased integration of renewable sources while providing clear benefits, such as high efficiency and simpler control. However, the protection of DC networks still remains a challenge due to strict time limits for fault interruption imposed by fast rising fault currents in DC systems, and absence of frequency and phasor information. This paper introduces a technique for fast detection and isolation of the faults in the DC microgrids without de-energizing the whole network. In the proposed algorithm, branch current measurements are sampled and Wavelet transform is applied to capture the characteristic changes in the current signals caused by network faults. The temporal variations in the relative wavelet energy within the frequency bands are acquired to construct the feature vector for classification. Artificial neural networks are used as the classifier as it provides a soft criterion for fault detection, featuring smart fault detection capability. The relatively fast calculation time of artificial neural networks makes it a good candidate for this application, due to the strict time restrictions inherited in DC fault isolation. To evaluate the performance, a comprehensive study on the proposed scheme is presented. The results demonstrate the effectiveness of the proposed scheme in terms of fast and reliable fault detection and inbuilt accurate fault localization capability.

Design and Implementation of Smart Fault Detection System for Industrial Power House using PLC and SCADA

Design and Implementation of Smart Fault Detection System for Industrial Power House using PLC and SCADA

A novel transmission line protection using DOST and SVM

This paper proposes a smart fault detection, classification and location (SFDCL) methodology for transmission systems with multi-generators using discrete orthogonal Stockwell transform (DOST). The methodology is based on synchronized current measurements from remote telemetry units (RTUs) installed at both ends of the transmission line. The energy coefficients extracted from the transient current signals due to occurrence of different types of faults using DOST are being utilized for real-time fault detection and classification. Support vector machine (SVM) has been deployed for locating the fault distance using the extracted coefficients. A comparative study is performed for establishing the superiority of SVM over other popular computational intelligence methods, such as adaptive neuro-fuzzy inference system (ANFIS) and artificial neural network (ANN), for more precise and reliable estimation of fault distance. The results corroborate the effectiveness of the suggested SFDCL algorithm for real-time transmission line fault detection, classification and localization.

A comprehensive study on the important faults in heat pump system during the warranty period

The heat pump market has become mature in many countries. There are millions of heat pumps installed worldwide. So any improvement in the installation, operation, and maintenance of heat pump systems can save a considerable amount of energy and cost, and reduce Green House Emissions to a large extent. The present study suggests a Smart Fault Detection and Diagnosis (SFDD) mechanism as the essential part of the next generation of heat pumps. A SFDD mechanism can minimize the installation and control errors, decrease the performance degradation during operation, avoid unnecessary visual inspections and components replacement, and reduce the maintenance cost and down-time of the system. To develop a SFDD mechanism, the first essential step is to obtain knowledge about the most common and expensive faults experienced by heat pumps. The heat pump manufacturers are one of the best sources to find out the most common and costliest faults occurring in heat pump systems during the first few years of their life. The present paper, as the first part of two, describes the results from a comprehensive study done on the most recent faults which were reported to some of the heat pump manufacturers in Sweden during the warranty period. The most common and the costliest faults in the Air/Air, Air/Water, Brine/Water, and exhaust air heat pumps are presented. Some of the faults such as faulty pressure switches or fans are only related to the heat pump unit, i.e. the thermodynamic cycle which facilitates the heat pumping cycle. Some of the common and expensive faults such as faulty shuttle or shunt valve are related to the faulty components in the heating systems. Generally, the results show that faults in Control and Electronics are almost the most common and costliest faults in all types of heat pumps. Faults in Control and Electronics include any fault related to control unit, electrical faults (such as short circuit, etc.), Printed Circuit Board (PCB), display, soft starter, overcurrent and motor protection relay, etc.

Fuzzy-Neuro Controler for Smart Fault Detection of a Beam

This paper addresses the fault detection of a cracked cantilever beam using a hybrid artificial intelligence technique. The hybrid technique used here uses a fuzzy-neuro controller. The fuzzy-neuro controller has two parts. The first part is comprised of the fuzzy controller, and the second part is comprised of the neural controller. The input parameters of the fuzzy controller are relative deviation of the first three natural frequencies and the relative values of the percentage deviation for the first three mode shapes. The output parameters of the fuzzy controller are initial relative crack depth and initial relative crack location. The input parameters of the neural segment of the fuzzy-neurocontroller are relative deviation of the first three natural frequencies and relative values of percentage deviation for the first three mode shapes, along with the initial outputs of the fuzzy controller. The output parameters of the fuzzy-neuro controller are final relative crack depth and final relative crack location. For deriving the fuzzy rules and training patterns of natural frequencies, mode shapes, crack depths and crack locations, theoretical expressions have been developed. Several fuzzy rules and training patterns for the fuzzy controller and neural controller of fuzzy-neuro controller are derived respectively. Experimental set-up has been developed for verifying the robustness of the fuzzy-neuro controller. The results of the developed fuzzy-neuro controller and experimental method are in good agreement.