• Journal of the KSME
• /
• v.53 no.7
• /
• pp.26-34
• /
• 2013

이 글에서는 최근 관심을 모으고 있는 고장예지 및 건전성관리(PHM: Prognostics and Health Management) 기술을 소개하고, 항공우주분야의 적용사례를 중심으로 PHM 기술을 어떻게 활용하고 있는지를 설명하고자 한다.

• Journal of the KSME
• /
• v.57 no.3
• /
• pp.34-37
• /
• 2017

이 글은 공학시스템 고장예지 및 건전성 관리 기술을 소개하고, 최근 주목받고 있는 인공지능기술인 딥러닝과 접목한 연구 사례를 제시한다.

• Journal of KSNVE
• /
• v.25 no.1
• /
• pp.22-28
• /
• 2015

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.42 no.4
• /
• pp.939-949
• /
• 2017

Prognostics and health management(PHM) is actively utilized by industry as an essential technology focusing on accurately monitoring the health state of a system and predicting the remaining useful life(RUL). An effective PHM is expected to reduce maintenance costs as well as improve safety of system by preventing failure in advance. With these advantages, PHM can be applied to the battery system which is a core element to provide electricity for devices with mobility, since battery faults could lead to operational downtime, performance degradation, and even catastrophic loss of human life by unexpected explosion due to non-linear characteristics of battery. In this paper we mainly review a recent progress on various models for predicting RUL of battery with high accuracy satisfying the given confidence interval level. Moreover, performance evaluation metrics for battery prognostics are presented in detail to show the strength of these metrics compared to the traditional ones used in the existing forecasting applications.

• Journal of the KSME
• /
• v.53 no.7
• /
• pp.40-43
• /
• 2013

이 글에서는 최근 관심을 모으고 있는 고장예지 및 건전성관리(PHM: Prognostics and Health Management) 기술이 자동차 전장부품에 어떻게 적용돼야 하는지에 대한 설명을 하고자 한다.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.19 no.9
• /
• pp.47-52
• /
• 2020

With the 4th industrial revolution, condition monitoring using machine learning techniques has become popular among researchers. An overload due to complex operations causes several irregularities in MOSFETs. This study investigated the acquired voltage to analyze the overcurrent effects on MOSFETs using a failure mode effect analysis (FMEA). The results indicated that the voltage pattern changes greatly when the current is beyond the threshold value. Several features were extracted from the collected voltage signals that indicate the health state of a switched-mode power supply (SMPS). Then, the data were reduced to a smaller sample space by using a principal component analysis (PCA). A robust machine learning algorithm, the support vector machine (SVM), was used to classify different health states of an SMPS, and the classification results are presented for different parameters. An SVM approach assisted by a PCA algorithm provides a strong fault diagnosis framework for an SMPS.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2017.07a
• /
• pp.9-10
• /
• 2017

본 논문에서 소음, 진동을 이용한 딥러닝 기반 기계 고장진단 임베디드 시스템을 제안하였다. 제안된 시스템은 기계로부터 취득된 소리와 진동을 바탕으로 학습한 DNN모델을 통해 실시간으로 기계 고장을 진단한다. 딥러닝 기술을 사용하여 학습에 따라 적용대상이 변경될 수 있도록 함으로써 특정 기계에 종속적이지 않고 가변적으로 다양한 기계에 대해 고장 예지 및 건전성 관리를 제공하도록 설계하였으며, 이를 증명하기 위해 액추에이터를 환풍기로 설정하여 정상상태와 4가지 비정상상태의 5가지상태를 학습하여 실험한 결과 93%의 정확도를 얻었다.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.40 no.8
• /
• pp.721-724
• /
• 2016

Because of the inherent uncertainties caused by the manufacturing process variations, future loading conditions, and incomplete damage models, the lifetimes of mechanical structures under field conditions are significantly different from the results obtained in the laboratories. In this study, a dual sensor was developed to prognosticate the fatigue failure of structures under these uncertain conditions, and its effectiveness was demonstrated on a rectangular columnar structure under repeated uni-axial loading. The dual sensor is a slightly weaker structure embedded in the target structure, so that failure occurs in the sensor earlier than in the target structure. From the signal differences in the strain gauges in the embedded dual sensor, it is possible to differentiate between the normal status and warning status, even under variable loads.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.9
• /
• pp.28-34
• /
• 2021

The Fourth Industrial Revolution has led to the development of drones for commercial and private applications. Therefore, the malfunction of drones has become a prominent problem. Failure mode and effect analysis was used in this study to analyze the primary cause of drone failure, and blade breakage was observed to have the highest frequency of failure. This was tested using a vibration sensor placed on drones along the breakage length of the blades. The data exhibited a significant increase in vibration within the drone body for blade fracture length. Principal component analysis was used to reduce the data dimension and classify the state with machine learning algorithms such as support vector machine, k-nearest neighbor, Gaussian naive Bayes, and random forest. The performance of machine learning was higher than 0.95 for the four algorithms in terms of accuracy, precision, recall, and f1-score. A follow-up study on failure prediction will be conducted based on the results of fault diagnosis.

• Journal of Aerospace System Engineering
• /
• v.13 no.6
• /
• pp.17-25
• /
• 2019

PHM (Prognostics and Health Management) of aircraft engines is applied to predict the remaining useful life before failure or the lifetime limit. There are two methods to establish a predictive model for this: The physics-based method and the data-driven method. The physics-based method is more accurate and requires less data, but its application is limited because there are few models available. In this study, the data-driven method is applied, in which a multi-layer perceptron based neural network algorithms is applied for the life prediction. The neural network is trained using the data sets virtually made by the C-MAPSS code developed by NASA. After training the model, it is applied to the test data sets, in which the confidence interval of the remaining useful life is predicted and validated by the actual value. The performance of proposed method is compared with previous studies, and the favorable accuracy is found.