• Journal of Electrical Engineering and Technology
• /
• 제10권4호
• /
• pp.1558-1565
• /
• 2015

In this paper, we propose a new method for detecting bearing faults using vibration signals. The proposed method is based on support vector machines (SVMs), which treat the harmonics of fault-related frequencies from vibration signals as fault indices. Using SVMs, the cross-validations are used for a training process, and a two-stage classification process is used for detecting bearing faults and their status. The proposed approach is applied to outer-race bearing fault detection in three-phase squirrel-cage induction motors. The experimental results show that the proposed method can effectively identify the bearing faults and their status, hence improving the accuracy of fault diagnosis.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.589-594
• /
• 1993

This paper presents a fault detection strategy that discriminates the faulty sensor and that detects the component fault using a bank of observers for the system in which sensor fault and component fault can occur simultaneously. Observers as many as the number of measurements are designed, and each observer uses measurements excluding sequentially one measurement, to estimate the state variables. The faulty sensor can be found out by comparing each state variable from different observer. Next, component fault can be detected by using measurements from the sensors excluding the faulty sensor. The suggested strategy is applied to a nonisothermal, series reaction with unknown reaction kinetics in a CSTR. This strategy is found out to perform well even in the case that the sensor and component fault occur simultaneously. Since each observer is designed to be independent of reaction kinetics, this strategy is not affected by the model uncertainty and nonlinearity of the reaction kinetics.

• 한국컴퓨터정보학회논문지
• /
• 제21권4호
• /
• pp.73-80
• /
• 2016

Most fault prediction models have class imbalance problems because training data usually contains much more non-fault class modules than fault class ones. This imbalanced distribution makes it difficult for the models to learn the minor class module data. Data imbalance is much higher when severity-based fault prediction is used. This is because high severity fault modules is a smaller subset of the fault modules. In this paper, we propose severity-based models to solve these problems using the three sampling methods, Resample, SpreadSubSample and SMOTE. Empirical results show that Resample method has typical over-fit problems, and SpreadSubSample method cannot enhance the prediction performance of the models. Unlike two methods, SMOTE method shows good performance in terms of AUC and FNR values. Especially J48 decision tree model using SMOTE outperforms other prediction models.

• 대한전기학회논문지:전력기술부문A
• /
• 제49권8호
• /
• pp.405-411
• /
• 2000

When faults occur in transmission lines, the classification of faults is very important. If the fault is HIF(High Impedance Fault), it cannot be detected or removed by conventional overcurrent relays (OCRs), and results in fire hazards and causes damages in electrical equipment or personal threat. The fast discrimination of fault needs to effective protection and treatment and is important problem for power system protection. This paper propolsed the fault detection and discrimination algorithm for LIFs(Low Impedance Faults) and HIFs(High Impedance Faults). This algorithm uses artificial neural networks and variation of 3-phase maximum currents per period while faults. A double lines-to-ground and line-to-line faults can be detected using Neural Network. Also, the other faults can be detected using the value of variation of maximum current. Test results show that the proposed algorithms discriminate LIFs and HIFs accurately within a half cycle.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 C
• /
• pp.1440-1443
• /
• 1999

This paper proposes fault detection method using a neural network & fuzzy logic on distribution lines. Fault on distribution lines is simulated using EMTP. The pattern of high impedance fault on pebbles, ground and short-circuit fault were take as the learning model. In this paper proposed fault detection method is evaluated on various conditions. The average values after analyzing fault current by FFT of even odd harmonics and fundamental rms were used for the neural network input. Test results were verified the validity of the proposed method

• Journal of Power Electronics
• /
• 제13권4호
• /
• pp.552-557
• /
• 2013

This paper proposed an analytical model for the distributed magnetic field analysis of interior permanent magnet-type blush-less direct current motors under the stator-turn fault condition using the winding function theory. Stator-turn faults cause significant changes in electric and magnetic characteristic. Therefore, many studies on stator-turn faults have been performed by simulation of the finite element method because of its non-linear characteristic. However, this is difficult to apply to on-line fault detection systems because the processing time of the finite element method is very long. Fault-tolerant control systems require diagnostic methods that have simple processing systems and can produce accurate information. Thus analytical modeling of a stator-turn fault has been performed using the winding function theory, and the distributed magnetic characteristics have been analyzed under the fault condition. The proposed analytical model was verified using the finite element method.

• 대한전기학회논문지:전력기술부문A
• /
• 제50권2호
• /
• pp.53-57
• /
• 2001

This paper presents the new methods of fault diagnosis through multiple alarm processing of protective relays and circuit breakers in power systems using probabilistic neural networks. In this paper, fault section detection neural network (FSDNN) for fault diagnosis is designed using the alarm information of relays or circuit breakers. In contrast to conventional methods, the proposed FSDNN determines the fault section directly and fast. To show the possibility of the proposed method, it is simulated through simulation panel for Sinyangsan substation system in KEPCO (Korea Electric Power Corporation) and the case studies show the effectiveness of the probabilistic neural network mehtod for the fault diagnosis.

• 전기학회논문지
• /
• 제67권9호
• /
• pp.1146-1151
• /
• 2018

Recently, researches on the intelligence of electric power facilities have been trying to apply artificial intelligence techniques as computer platforms have improved. In particular, faults occurring in substation should be able to quickly identify possible faults and minimize power fault recovery time. This paper presents fault location technique for 154kV substation using neural network. We constructed a training matrix based on the operating conditions of the circuit breaker and IED to identify the fault location of each component of the target 154kV substation, such as line, bus, and transformer. After performing the training to identify the fault location by the neural network using Weka software, the performance of fault location discrimination of the designed neural network was confirmed.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2005년도 하계학술대회 논문집 Vol.6
• /
• pp.225-226
• /
• 2005

In this Study, we have acquired 5-simulation Fault types Signals of high voltage Motor stator winding using epoxy/mica coupler. In order to know stator winding fault type using fault signals, we have performed feature extraction to apply wavelet transform technique. we have obtained skewness and kurtosis as statistical parameters of fault signal pattern from non deterioration state winding. We have know that 5 fault signals types have done an exponential function pattern shape but individually fault a class widely was different each other a signal waveform of pattern.

• Journal of information and communication convergence engineering
• /
• 제19권3호
• /
• pp.175-179
• /
• 2021

Enhancing the reliability of the system is important for recent system-on-chip (SoC) designs. This importance has led to studies on fault diagnosis and tolerance. Fault-injection (FI) techniques are widely used to measure the fault-tolerance capabilities of resilient systems. FI techniques suffer from limitations in relation to environmental conditions and system features. Moreover, a hardware-based FI can cause permanent damage to the target system, because the actual circuit cannot be restored. Accordingly, we propose a simulation-based FI framework based on the Verilog Procedural Interface for measuring the failure rates of SoCs caused by soft errors. We execute five benchmark programs using an ARM Cortex M0 processor and inject soft errors using the proposed framework. The experiment has a 95% confidence level with a ±2.53% error, and confirms the reliability and feasibility of using proposed framework for fault analysis in SoCs.