• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2007년도 제28회 춘계학술대회논문집
• /
• pp.245-249
• /
• 2007

본 연구에서는 모델 기반(Model-Based) 성능진단에 신경회로망을 적용하였고, SIMULINK를 이용하여 PW206C 터보축 엔진의 모델링을 수행하였다. 비행 고도, 비행 마하수, 가스발생기 회전수에 따른 다양한 운용영역의 성능데이터를 base로 하여 압축기, 압축기터빈, 동력터빈의 성능 저하에 대한 학습 데이터를 획득하고 역전파(Back Propagation Network)를 이용하여 훈련 하였다. 설계점 및 탈설계 영역에서 압축기, 압축기터빈, 동력터빈의 단일 손상 탐지를 수행한 결과 손상된 구성품을 잘 탐지함을 확인할 수 있었다.

• 한국추진공학회지
• /
• 제11권3호
• /
• pp.29-34
• /
• 2007

본 연구에서는 모델 기반(Model-Based) 성능진단에 신경회로망을 적용하였고, SIMULINK를 이용하여 PW206C 터보축 엔진의 모델링을 수행하였다. 비행 고도, 비행 마하수, 가스발생기 회전수에 따른 다양한 운용영역의 성능데이터를 base로 하여 압축기, 압축기터빈, 동력터빈의 성능 저하에 대한 학습데이터를 획득하고 역전파(Back Propagation Network)를 이용하여 훈련하였다. 설계점 및 탈설계 영역에서 압축기, 압축기터빈, 동력터빈의 단일 손상 탐지를 수행한 결과 손상된 구성품을 비교적 잘 탐지함을 확인할 수 있었다.

• 설비공학논문집
• /
• 제12권7호
• /
• pp.688-696
• /
• 2000

A scheme for fault detection on the subsystem level is presented. The method uses analytical redundancy and consists in generating residuals by comparing each measurement with an estimate computed from the reference models. In this study regression neural network models are used as reference models. The regression neural network is memory-based feed forward network that provides estimates of continuous variables. The simulation result demonstrated that the proposed method can effectively detect faults in an air handling unit(AHU). The results show that the regression models are accurate and reliable estimators of the highly nonlinear and complex AHU.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2008년도 춘계학술대회 논문집 전기설비전문위원
• /
• pp.171-173
• /
• 2008

This paper presents stochastic methodology based fault diction and diagnosis algorithm for induction motor systems. First, we construct probability distribution model from healthy motors and then probability distribution for faulty motors is recursively calculated by means of the proposed probability estimation. We measure motor current with hall sensors as system state. The estimated probability is compared to the model to generate a residue signal which is utilized for fault detection and diagnosis, that is, where a fault is occurred. We carry out real-time induction motor experiment to evaluate efficiency and reliability of the proposed approach.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2008년도 제39회 하계학술대회
• /
• pp.1503-1504
• /
• 2008

This paper presents stochastic methodology based fault diction and diagnosis algorithm for induction motor systems. First, we construct probability distribution model from healthy motors and then probability distribution for faulty motors is recursively calculated by means of the proposed probability estimation. We measure motor current with hall sensors as system state. The estimated probability is compared to the model to generate a residue signal which is utilized for fault detection and diagnosis, that is, where a fault is occurred. We carry out real-time induction motor experiment to evaluate efficiency and reliability of the proposed approach.

• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 2001년도 The Seoul International Simulation Conference
• /
• pp.301-305
• /
• 2001

A novel approach based on parameters estimation is presented far fault detection and diagnosis in sensors. Based on known precise parameter of normal working sensors system model is built from real laboratory inputs-outputs data, sequentially residual serial is obtained. Where decision-making rule of detection the fault is given via the use of beys theory, whilst a filter least-square computative algorithm for estimating fault parameters is given. The algorithm is a fast and accurate to calculate value of sensors faults when system model contains noise and sensors outputs contain measured noise. The method can solve both gain type and bias type fault in sensors. Simulated numerical example is included to demonstrate the use of the proposed approaches.

• 디지털융복합연구
• /
• 제11권11호
• /
• pp.335-342
• /
• 2013

본 연구에서는 소프트웨어 제품 테스팅 과정에서 관측고장시간에 근거한 로그형 결함 발생률을 고려한 소프트웨어 신뢰성 비용 모형에 대하여 연구 하였다. 신뢰성 분야에서 많이 사용되는 Goel-Okumoto모형을 이용한 새로운 로그 형 결함 확률을 반영한 문제를 제시하였다. 수명분포는 유한고장 비동질적인 포아송과정을 이용하고 모수 추정법은 최우 추정법을 이용 하였다. 따라서 본 논문에서는 로그형 결함 발생률을 고려한 소프트웨어 비용모형 분석을 위하여 소프트웨어 고장 시간간격 자료를 적용하여 비교 분석하였다. 이 연구를 통하여 소프트웨어 개발자들은 방출최적시기를 파악 하는데 어느 정도 도움을 줄 수 있을 것으로 사료 된다.

• 산업공학
• /
• 제22권2호
• /
• pp.126-134
• /
• 2009

As product quality and yield are essential factors in semiconductor manufacturing, monitoring the main manufacturing steps is a critical task. For the purpose, FDC(Fault detection and classification) is used for diagnosing fault states in the processes by monitoring data stream collected by equipment sensors. This paper proposes an FDC model based on decision tree which provides if-then classification rules for causal analysis of the processing results. Unlike previous decision tree approaches, we reflect the structural aspect of the data stream to FDC. For this, we segment the data stream into multiple subregions, define structural features for each subregion, and select the features which have high relevance to results of the process and low redundancy to other features. As the result, we can construct simple, but highly accurate FDC model. Experiments using the data stream collected from etching process show that the proposed method is able to classify normal/abnormal states with high accuracy.

• Smart Structures and Systems
• /
• 제23권4호
• /
• pp.393-403
• /
• 2019

Applying more features gives us better accuracy in modeling; however, increasing the inputs causes the curse of dimensions. In this paper, a new structure has been proposed for fault detecting and identifying (FDI) of high-dimensional systems. This structure consist of two structure. The first part includes Auto-Encoders (AE) as Deep Neural Networks (DNNs) to produce feature engineering process and summarize the features. The second part consists of the Local Model Networks (LMNs) with LOcally LInear MOdel Tree (LOLIMOT) algorithm to model outputs (multiple models). The fault detection is based on these multiple models. Hence the residuals generated by comparing the system output and multiple models have been used to alarm the faults. To show the effectiveness of the proposed structure, it is tested on single-shaft industrial gas turbine prototype model. Finally, a brief comparison between the simulated results and several related works is presented and the well performance of the proposed structure has been illustrated.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제52권10호
• /
• pp.575-582
• /
• 2003

A new FDOs (fault diagnostic observers) and the residual generation schemes using the FDOs are suggested for the process fault detection and isolation of linear (control) systems. The design method of the FDO is described, first, for the full measurement systems. Then it is extended for the systems with unmeasurable state variables. An unknown input observer is proposed and applied for the extension. The size of the observer bank may be the smallest, specially in full measurement systems, because the order of the proposed FDO is very low. In spite of the simplicity, the scheme provides the same information for the detection and isolation of the anticipated faults as the conventional multiple observer based schemes. The residuals may be structured so that fault isolation can be performed by pre-selected logic. An FDIS using the proposed scheme is constructed for the model of the four-tank system. Simulation results show the practical feasibility of the proposed scheme.