• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2011.11a
• /
• pp.94-95
• /
• 2011

선박 엔진 크랭크 케이스 내부의 기계 결함으로 발생하는 윤활 오일 미스트 발생은 폭발뿐만 아니라 급격한 온도와 압력상승으로 2차 폭발을 야기하는 등 큰 피해를 가져와 현재 선박용 엔진의 크랭크 케이스에 오일미스트 검출장치의 설치를 의무화 하고 있는 추세이다. 본 논문에서는 이러한 피해를 줄이기 위하여 광 산란 방식을 이용한 오일 미스트 검출장치를 구현하였으며, 구현된 오일미스트 검출 장치의 오일미스검출 센서 센싱 성능은 전기적, 광학적 그 외 환경적 요인으로 인하여 출력값이 일정하지 못해 각각의 오일미스트 검출 센서를 보정해야 한다. 본 논문에서는 크랭크케이스마다 장착되는 오일미스트 검출 센서의 오차를 최소화하기 위하여 정밀하게 Calibration된 Reference Data를 기반으로 최소자승법을 적용하여 센서를 Calibration하였으며, 그 결과 각각의 오일미스 검출센서의 오차를 기존 방법과 비교하여 최소화할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.04a
• /
• pp.749-752
• /
• 2014

본 논문에서는 저속으로 동작하는 구름 베어링의 다중 결함 조기 검출을 위해 결함 특징 추출, 효과적인 특징 선택, 선택된 특징을 이용한 결함 분류의 세 단계로 구성된 결함 진단 기법을 제안한다. 1단계에서 이산 웨이블릿 변환을 이용하여 미세성분으로부터 통계적 결함 특징을 추출하고, DET(distance evaluation technique)를 이용하여 추출한 결함 특징 가운데 베어링 다중 결함 검출에 효과적인 특징을 선택한다. 마지막으로 선택된 특징을 k-NN(k-Nearest Neighbors) 분류기 입력으로 사용함으로써 결함을 진단한다. 본 논문에서는 제안한 결함 진단 기법의 성능을 분류 정확도 측면에서 평가한 결과 95.14%의 높은 분류 정확도를 보였다.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.19 no.8
• /
• pp.131-137
• /
• 2005

The purpose of this paper is to evaluate the axial magnetic flux measurement could be used as a tool of the condition monitoring system for the induction motor and to develope the diagnostic algerian for the electric motors. The magnetic leakage flux signal is captured by the flux coil located at the end of motor without the disturbance of the operation. And the signal is analyzed both time and frequency bases to detect the failure of the motor. Specific signature can be described in time and frequency domain for each faults of the motor. The spectrum of the signal was found more useful for the monitoring purpose. The supply voltage imbalance and tin to turn failure of the stator winding could be detected by analysing the specific sidebands of the axial flux and sideband of the rotor bar pass frequency with the high resolution spectrum. The goal of this study verity that the axial flux measurement for the induction motor is a powerful tool for the diagnostic method and develope the algorithm to detect the fault.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.334-336
• /
• 2001

반도체 제조와 같은 대량 생산 시스템에서 제품 검사는 매우 중요란 단계 중의 하나이다. 반도체 제조 공정 내에서의 시각 검사는 현재 사람의 육안에 주로 의존하고 있으나, 회로가 점점 복잡해지고 작아지는 추세에 비추어 볼 때 사람에 의한 시각 검사는 한계에 이를 것으로 보인다. 본 연구에서는 웨이퍼상의 결함을 자동으로 검출하고 검출된 길함을 분류하는 자동시각검사 시스템을 설계 구현하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.2
• /
• pp.313-318
• /
• 2012

This paper presents an automatic system to detect variety of faults from fine pitch COF(chip-on-film) which is less than $30{\mu}m$. Developed system contains circuits and technique to detect fast various faults such as hard open, hard short, soft open and soft short from fine pattern. Basic principle for fault detection is to monitor fine differential voltage from pattern resistance differences between fault-free and faulty cases. The technique uses also radio frequency resonator arrays for easy detection to amplify fine differential voltage. We anticipate that proposed system is to be an alternative for conventional COF test systems since it can fast and accurately detect variety of faults from fine pattern COF test process.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.10a
• /
• pp.153-153
• /
• 2003

• Journal of Korea Water Resources Association
• /
• v.55 no.spc1
• /
• pp.1177-1185
• /
• 2022

The water and sewage system is an infrastructure that provides safe and clean water to people. In particular, since the water and sewage pipelines are buried underground, it is very difficult to detect system defects. For this reason, the diagnosis of pipelines is limited to post-defect detection, such as system diagnosis based on the images taken after taking pictures and videos with cameras and drones inside the pipelines. Therefore, real-time detection technology of pipelines is required. Recently, pipeline diagnosis technology using advanced equipment and artificial intelligence techniques is being developed, but AI-based defect detection technology requires a variety of learning data because the types and numbers of defect data affect the detection performance. Therefore, in this study, various defect scenarios are implemented using 3D printing model to improve the detection performance when detecting defects in pipelines. Afterwards, the collected images are performed to pre-processing such as classification according to the degree of risk and labeling of objects, and real-time defect detection is performed. The proposed technique can provide real-time feedback in the pipeline defect detection process, and it would be minimizing the possibility of missing diagnoses and improve the existing water and sewerage pipe diagnosis processing capability.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.30 no.1
• /
• pp.31-35
• /
• 2010

This paper presents the development of automatic defect detection program using 3D computed tomography image of small solid rocker motor. We applied the neighbor pixel comparison algorithm with beam hardening correction for the recognition of defect. We made the artificial defect specimen in order to decide a standard CT value of defect. The program was tested with 150 small solid rocket motors and it could detect the disbond, crack, foreign material and void. The program showed more reliable and faster results than human inspector's interpretation.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.18 no.6
• /
• pp.445-454
• /
• 1998

In order to establish a technical concept for the detection of defects in weldments in thin steel plate, an experimental and theoretical investigation was carried out for artificial defects in a steel plate having a thickness of 2.4mm by using the guided wave technique. In particular the goal was to find the most effective testing parameters paying attention to the relationship between the excitation frequency by a tone burst system and various incident angles. It was found that the test conditions that worked best was for a frequency of 840kHz and an incident angle of 30 or 85 degrees, most of the defects were detected with these conditions. Also, it was clear that a guided wave mode generated under an incident angle of 30 degrees was a symmetric mode, So, and that of 85 degrees corresponded to an antisymmetric mode, Ao. By using the two modes, most of all of the defects could be detected. Furthermore, it was shown that the antisymmetric mode was more sensitive to defects near the surface than the symmetric mode. Theoretical predictions confirmed this sensitivity improvement with Ao for surface defects because of wave structure variation and energy concentration near the surface.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.4
• /
• pp.577-586
• /
• 2022

Detection and classification of steel surface defects are critical for product quality control in the steel industry. However, due to its low accuracy and slow speed, the traditional approach cannot be effectively used in a production line. The current, widely used algorithm (based on deep learning) has an accuracy problem, and there are still rooms for development. This paper proposes a method of steel surface defect detection combining EfficientNetV2 for image classification and YOLOv5 as an object detector. Shorter training time and high accuracy are advantages of this model. Firstly, the image input into EfficientNetV2 model classifies defect classes and predicts probability of having defects. If the probability of having a defect is less than 0.25, the algorithm directly recognizes that the sample has no defects. Otherwise, the samples are further input into YOLOv5 to accomplish the defect detection process on the metal surface. Experiments show that proposed model has good performance on the NEU dataset with an accuracy of 98.3%. Simultaneously, the average training speed is shorter than other models.