• Transactions on Electrical and Electronic Materials
• /
• 제6권6호
• /
• pp.262-271
• /
• 2005

In order to achieve timely and accurate fault detection of plasma etching process, neural network based time series modeling has been applied to reactive ion etching (RIE) using two different in-situ plasma-monitoring sensors called optical emission spectroscopy (OES) and residual gas analyzer (RGA). Four different subsystems of RIE (such as RF power, chamber pressure, and two gas flows) were considered as potential sources of fault, and multiple degrees of faults were tested. OES and RGA data were simultaneously collected while the etching of benzocyclobutene (BCB) in a $SF_6/O_2$ plasma was taking place. To simulate established TSNNs as incipient fault detectors, each TSNN was trained to learn the parameters at t, t+T, ... , and t+4T. This prediction scheme could effectively compensate run-time-delay (RTD) caused by data preprocessing and computation. Satisfying results are presented in this paper, and it turned out that OES is more sensitive to RF power and RGA is to chamber pressure and gas flows. Therefore, the combination of these two sensors is recommended for better fault detection, and they show a potential to the applications of not only incipient fault detection but also incipient real-time diagnosis.

• 한국태양에너지학회 논문집
• /
• 제33권4호
• /
• pp.89-93
• /
• 2013

In recent years, many investigations about photovoltaic power systems have been significantly carried out in the fields of renewable power energy. Such research area generally includes developments of highly efficient solar cells, advanced power conversion systems, and smart monitoring systems. A generic objective of fault detection and diagnosis techniques is to timely recognize unexpected faulty of dynamic systems so that economic demage occurred by such faulty is decreased by means of engineering techniques. This paper presents a novel fault detection approach for photovoltaic power arrays which are electrically connected in series and parallels. In the proposed fault detection scheme, we first measure all of photovoltaic modules located in each array by using electronic sense systems and then compare each measurement in turn to detect location of fault module through statistic computation algorithm. We accomplish real-time experiments to demonstrate our proposed fault detection methodology by using a test-bed system including two 20 watt photovoltaic modules.

• Geomechanics and Engineering
• /
• 제23권2호
• /
• pp.127-137
• /
• 2020

To study the reverse fault formation process and the stress evolution feature, a simulation test system of reverse fault formation is developed based on the analysis of reverse fault formation mechanism. The system mainly consists of simulation laboratory module, operation console and horizontal loading control system, and data monitoring system. It can represent the fault formation process, induce fault crack initiation and simulate faults of different throws. Simulation tests on reverse fault formation process are conducted by using the simulation test system: horizontal loading is added to one side of the model. the bottom rock layer cracks under the effect of the induction device. The crack dip angle is about 29°. A reverse fault is formed with the expansion of the crack dip angle towards the upper right along the fracture surface and the slippage of the hanging wall over the foot wall. Its formation process unfolds five stages: compressive deformation of rock, local crack initiation, reverse fault penetration, slippage of the hanging wall over the foot wall and compaction of fault plane. There is residual structural stress inside rock after fault formation. The study methods and results have guiding and referential significance for further study on reverse fault formation mechanism and rock stress evolution.

• 한국신재생에너지학회:학술대회논문집
• /
• 한국신재생에너지학회 2005년도 춘계학술대회
• /
• pp.388-391
• /
• 2005

In this paper, Cell voltage monitoring method is studied for fault detection of PEMFC(Proton Exchange Membrane Fuel Cell) for FCEV(fuel cell electric vehicle). To measuring several hundred of cells in fuel cell stack, The demanded feature of hardware and software is studied and several types are analysed. Finally, $3.26\%$ maximum measuring error is acquired and verified experimentally.

• 한국지능시스템학회논문지
• /
• 제24권1호
• /
• pp.78-83
• /
• 2014

최근 신재생 에너지원으로서의 선두주자인 풍력발전은 다수의 풍력발전 회사들로 하여금 모니터링 및 고장진단 시스템의 개발을 가속화시키고 있다. 이러한 모니터링 및 진단시스템은 조기의 고장검출을 통해 고장이 발생되었을 경우 발생되는 고가의 수리비용을 미연에 방지할 수 있게 한다. 일반적으로 풍력발전과 관련된 고장진단 시스템은 진동신호 및 신호분석기법에 기반하고 있다. 이에 본 연구에서는 풍력발전 시스템에서 자주 발생되고 있는 질량 불평형 및 축 정렬 불량 등과 같은 기계적인 고장을 효율적으로 진단할 수 있는 시스템을 제안하고자 한다. 본 연구에서 제안된 지능화된 고장진단 알고리즘은 인공신경망기법과 웨이블렛 변환을 이용한 것으로 (주)가온솔루션에서 개발한 풍력발전용 기계적 고장발생 장치에 적용 실험을 통해 제안된 진단기법의 유용성을 확인하고자 하였다.

• 대한임베디드공학회논문지
• /
• 제19권3호
• /
• pp.151-158
• /
• 2024

Although on-device artificial intelligence (AI) has gained attention to diagnosing machine faults in real time, most previous studies did not consider the model retraining and redeployment processes that must be performed in real-world industrial environments. Our study addresses this challenge by proposing an on-device AI-based real-time machine fault diagnosis system that utilizes continual learning. Our proposed system includes a lightweight convolutional neural network (CNN) model, a continual learning algorithm, and a real-time monitoring service. First, we developed a lightweight 1D CNN model to reduce the cost of model deployment and enable real-time inference on the target edge device with limited computing resources. We then compared the performance of five continual learning algorithms with three public bearing fault datasets and selected the most effective algorithm for our system. Finally, we implemented a real-time monitoring service using an open-source data visualization framework. In the performance comparison results between continual learning algorithms, we found that the replay-based algorithms outperformed the regularization-based algorithms, and the experience replay (ER) algorithm had the best diagnostic accuracy. We further tuned the number and length of data samples used for a memory buffer of the ER algorithm to maximize its performance. We confirmed that the performance of the ER algorithm becomes higher when a longer data length is used. Consequently, the proposed system showed an accuracy of 98.7%, while only 16.5% of the previous data was stored in memory buffer. Our lightweight CNN model was also able to diagnose a fault type of one data sample within 3.76 ms on the Raspberry Pi 4B device.

• 한국소음진동공학회논문집
• /
• 제20권11호
• /
• pp.1009-1017
• /
• 2010

As rotating machines play an important role in industrial applications such as aeronautical, naval and automotive industries, many researchers have developed various condition monitoring system and fault diagnosis system by applying various techniques such as signal processing and pattern recognition. Recently, fault diagnosis systems using artificial neural network have been proposed. For effective fault diagnosis, this paper used MLP(multi-layer perceptron) network which is widely used in pattern classification. Since using obtained signals without preprocessing as inputs of neural network can decrease performance of fault classification, it is very important to extract significant features of captured signals and to apply suitable features into diagnosis system according to the kinds of obtained signals. Therefore, this paper proposes the decision method of the proper feature vectors about each fault signal for neural-network-based fault diagnosis system. We applied LPC coefficients, maximum magnitudes of each spectral section in FFT and RMS(root mean square) and variance of wavelet coefficients as feature vectors and selected appropriate feature vectors as comparing error ratios of fault diagnosis for sound, vibration and current fault signals. From experiment results, LPC coefficients and maximum magnitudes of each spectral section showed 100 % diagnosis ratios for each fault and the method using wavelet coefficients had noise-robust characteristic.

• Journal of Platform Technology
• /
• 제6권3호
• /
• pp.47-54
• /
• 2018

하드웨어 산업에서의 예측 분석은 생산설비의 고장을 방지하기 위해 적절한 시점에 유지보수를 수행할 수 있고 관리비용을 절감할 수 있다. 이는 고장원인분석의 자동화를 통해 보다 효율적이고 과학적인 유지보수를 수행할 수 있도록 도와준다. 그중에서도 예측 관리는 정보 기술을 활용하여 설비 상태의 수집, 분석, 과학적 데이터 관리를 통해 예측 모델을 구성하며, 이를 바탕으로 이상상태를 파악하고 개선함으로써 이상상태가 발생하는 것을 사전에 예방하는 것을 목적으로 한다. 본 연구에서는 Relex 도구를 통해 결함트리(Fault Tree)를 만들고 하드웨어들의 에러코드를 분석하여 안전성을 연구했다.

• 한국소음진동공학회논문집
• /
• 제20권2호
• /
• pp.185-190
• /
• 2010

Condition monitoring(CM) is a method based on non-destructive test(NDT). So, recently many kind of NDT were applied for CM. Acoustic emission(AE) is widely used for the early detection of faults in rotating machinery in these days because of high sensitivity than common accelerometers and detectable low energy vibration signals. And crack is considered one of severe fault in the rotating machine. Therefore, in this paper, study on early detection using AE has been accomplished for the crack of the low-speed shaft. There is a seeded initial crack on the shaft then the AE signal had been measured with low-speed rotation as the applied load condition. The signal detected from crack in rotating machine was detected by the AE transducer then the trend of crack growth had found out by using some of feature values such as peak value, skewness, kurtosis, crest factor, frequency center value(FC), variance frequency value(VF) and so on.

• Smart Structures and Systems
• /
• 제29권6호
• /
• pp.785-790
• /
• 2022

Condition monitoring of permanent magnet synchronous motors (PMSMs) and detecting faults such as eccentricity and demagnetization are essential for ensuring system reliability. Motor current signal analysis is the most commonly used precursor for detecting faults in the PMSM drive system. However, the current signature responds sensitively to the load and temperature of the motor, thereby making it difficult to monitor faults in real- applications. Therefore, in this study, a condition monitoring methodology that detects motor faults, including their classification with standstill conditions, is proposed. The objective is to detect and classify faults of PMSMs by using programmable inverter without additional sensors and systems for detection. Both DC and AC were applied through the d-axis of a three-phase motor, and the change in incremental inductance was investigated to detect and classify faults. Simulation with finite element analysis and experiments were performed on PMSMs in healthy conditions as well as with eccentricity and demagnetization faults. Based on the results obtained from experiments, the proposed method was confirmed to detect and classify types of faults, including their severity.