• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2022.11a
• /
• pp.265-267
• /
• 2022

선박 기관시스템이 효율적이고 안이정적인 운용을 위해서는 실시간 상태 모니터링 기반의 이상탐지, 고장진단 더 나아가 고장예측에 따른 대응조치를 할 수 있는 기술이 필요하며 이를 상태기반 유지관리(Condition Based Maintenance, CBM)이라 지칭한다. 해당 기술을 개발 및 확보하기 위해서는 가장 우선적으로 기관시스템에 대한 다양한 고장 데이터가 확보되어야 하며 이후, 확보된 데이터에 대한 특징추출 등 전처리 알고리즘, 고장 진단 및 예측 알고리즘 등을 개발하여야 한다. 본 연구에서는 선박 추진용 엔진 및 발전기 엔진에 대한 상태기반 유지관리 기술의 개발현황과 향후 지속적인 연구 추진방향을 소개하고자 한다.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.3
• /
• pp.500-506
• /
• 2017

In 2014, ISO 55000s has been enacted and the power plant asset management is becoming a hot issue for all over the world. The asset management system is being developed as a combination of CBM(Condition Based Maintenance) and RCM(Reliability Centered Maintenance). Therefore, the research on the calculation of the failure rate which is the most basic index of RCM is actively carried out. The failure rate calculation has been going on for a long time, and the most widely used probability distribution is the Weibull distribution. In the Weibull distribution, the failure rate function is determined in three types according to the value of the shape parameter. However, the Weibull distribution has a limitation that it is difficult to apply it when the trend of failure rate changes-such as bathtub curves. In this paper, the failure rate is calculated using the mixture Weibull distribution which can appropriately express the change of the shape of the failure rate. Based on these results, we propose the necessity and validity of applying mixture Weibull distribution.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.2
• /
• pp.133-142
• /
• 2023

The large process plant is currently implementing predictive maintenance technology to transition from the traditional Time-Based Maintenance (TBM) approach to the Condition-Based Maintenance (CBM) approach in order to improve equipment maintenance and productivity. The traditional techniques for predictive maintenance involved managing upper/lower thresholds (Set-Point) of equipment signals or identifying anomalies through control charts. Recently, with the development of techniques for big analysis, machine learning-based AAKR (Auto-Associative Kernel Regression) and deep learning-based VAE (Variation Auto-Encoder) techniques are being actively applied for predictive maintenance. However, this predictive maintenance techniques is only effective during steady-state operation of plant equipment, and it is difficult to apply them during start-up and shutdown periods when rises or falls. In addition, unlike processes such as nuclear and thermal power plants, which operate for hundreds of days after a single start-up, because the pumped power plant involves repeated start-ups and shutdowns 4-5 times a day, it is needed the prediction and alarm algorithm suitable for its characteristics. In this study, we aim to propose an approach to apply the optimal predictive alarm algorithm that is suitable for the characteristics of Pumped Storage Power Plant(PSPP) facilities to the system by analyzing the predictive maintenance techniques used in existing nuclear and coal power plants.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.28 no.4
• /
• pp.191-201
• /
• 2016

A stochastic process has been used to develop a condition-based model for preventive maintenance of armor units of rubble-mound breakwaters that can make a decision the optimal interval at which some repair actions should be performed under the perfect maintenance. The proposed cost model in this paper based on renewal reward process can take account of the interest rate, also consider the unplanned maintenance cost which has been treated like a constant in the previous studies to be a time-dependent random variable. A function for the unplanned maintenance cost has been mathematically proposed so that the cumulative damage, serviceability limit and importance of structure can be taken into account, by which a age-based maintenance can be extended to a condition-based maintenance straightforwardly. The coefficients involved in the function can also be properly estimated using a method expressed in this paper. Two stochastic processes, Wiener process and gamma process have been applied to armor stones of rubble-mound breakwaters. By evaluating the expected total cost rate as a function of time for various serviceability limits, interest rates and importances of structure, the optimal period of preventive maintenance can easily determined through the minimization of the expected total cost rate. For a fixed serviceability limit, it shows that the optimal period has been delayed while the interest rate increases, so that the expected total cost rate has become lower. In addition, the gamma process tends to estimate the optimal period more conservatively than the Wiener process. Finally, it is found that the more crucial the level of importance of structure becomes, the more often preventive maintenances should be carried out.

• Journal of Korean Society for Quality Management
• /
• v.51 no.1
• /
• pp.119-129
• /
• 2023

Purpose: This study proposes a classification model for implementing condition-based maintenance (CBM) by monitoring the real-time status of a machine using acceleration sensor data collected from a vehicle. Methods: The classification model's performance was improved by applying Fourier transform to convert the acceleration sensor data from the time domain to the frequency domain. Additionally, the Generative Adversarial Network (GAN) algorithm was used to augment images and further enhance the classification model's performance. Results: Experimental results demonstrate that the GAN algorithm can effectively serve as an image augmentation technique to enhance the performance of the classification model. Consequently, the proposed approach yielded a significant improvement in the classification model's accuracy. Conclusion: While this study focused on the effectiveness of the GAN algorithm as an image augmentation method, further research is necessary to compare its performance with other image augmentation techniques. Additionally, it is essential to consider the potential for performance degradation due to class imbalance and conduct follow-up studies to address this issue.

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.1200-1201
• /
• 2015

고전압 전력설비 진단은 기기의 열화 상태를 측정하여 이상 사고를 미리 예측하여 방지하는 것을 목적으로 실시한다. 고전압 전력설비의 유지관리 방안은 일정 시간 경과후 보수하는 개념(TBM: Time Based Maintenance) 이후 설비의 상태를 진단하여 유지보수 방안을 결정하는 개념(CBM: Condition Based Maintenance)으로 진보해 감에 따라 전력설비의 상태진단 기술의 중요성은 증대될 전망이다. 고전압 전력설비의 절연진단은 직류시험(절연저항, PI)과 교류시험($tan{\delta}$, PD)이 실시되며 과거 진단 데이터의 추세분석을 통한 정확한 상태진단이 요구되고 있다. 고압 유도 전동기 절연진단 데이터 관리 전산화 모델은 고전압 전력기기(발전기, 변압기, 전동기, 케이블 등)의 절연진단 및 유지보수 이력에 관한 자료들을 저장, 조회 및 검색을 하기 위한 데이터베이스를 구축하고 구축된 데이터를 활용하여 과거 이력조회, 추이분석, 진단 데이터의 분석기법을 통한 전력기기의 상태평가로 합리적인 개 대체 의사결정을 지원한다. 또한, 유입식 변압기의 절연유 가스분석 알고리즘을 전산화 하여 10종 가연성 가스에 따른 Gas Pattern 평가로 고장 원인, 현상 및 조치 등에 대한 출력이 가능한 프로그램의 개발로 고전압 전력설비 진단기술과 IT기술의 융 복합 기술로서 고전압 전력설비의 유지관리 기술을 한 차원 더 진보시킬 것으로 판단된다.

• Journal of the Korean Institute of Gas
• /
• v.11 no.4
• /
• pp.64-72
• /
• 2007

Under the AE methods, the valid condition analysis and evaluation the leak etc, resulted by the AE signal pattern on the bottom plate of ground tank at full. In next more, the gradient of accumulation amplitude distribution analysis and comparison the energy, count, and duration time that noise of EMI signal were removed. EMI signal showed height-energy, count, and duration time, it also appeared great gradient of accumulation distribution. Then, with the pure remaining AE signals cluster analysis and location. It would possibly assume of damage with corrosion. Total cluster 20 and energy showed between the maximum 11,990 and 8,565 which is much lower than above figure and event number showed from 8 to 5. Even when it difficult to certify damage by open, as it is raised higher height-sensitivity and threshold by 60 dB. It would possibly presume of location source more accurately.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.1
• /
• pp.127-134
• /
• 2021

In this study, an anomaly detection (AD) algorithm was implemented to detect the failure of a marine air compressor. A lab-scale experiment was designed to produce fault datasets (time-series electric current measurements) for 10 failure modes of the air compressor. The results demonstrated that the temporal pattern of the datasets showed periodicity with a different period, depending on the failure mode. An AD model with a convolutional autoencoder was developed and trained based on a normal operation dataset. The reconstruction error was used as the threshold for AD. The reconstruction error was noted to be dependent on the AD model and hyperparameter tuning. The AD model was applied to the synthetic dataset, which comprised both normal and abnormal conditions of the air compressor for validation. The AD model exhibited good detection performance on anomalies showing periodicity but poor performance on anomalies resulting from subtle load changes in the motor.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.29 no.7
• /
• pp.429-434
• /
• 2016

This paper deals with the characteristics of partial discharge (PD) for the purpose of a condition based maintenance (CBM) of gas insulated switchgears (GIS) in power equipment. Four types of electrode systems such as a protrusion on enclosure (POE), a particle on spacer (POS), a free particle (FP) and a Floating were designed and fabricated. PD pulses were measured using UHF sensor with a frequency range of 300 MHz~1.4 GHz and a DAQ with a sampling rate of 250 MS/s. Discharge inception voltage (DIV), discharge extinction voltage (DEV), and phase resolved partial discharge (PRPD) were analyzed depending on electrode systems. The average DIV in the POS was 28.8 kV. It was about 1.7 times higher than that in the FP, which was the lowest value of 17.2 kV. The FP shuffled and jumped at the applied voltage of 23.5 kV. Over 95% of PD pulses in the POE were generated in the negative polarity ($181^{\circ}{\sim}360^{\circ}$) of applied voltage. The results showed the phase (${\Phi}$)-magnitude (dBm) of PD pulses by UHF sensor, a cluster was formed separately depending on electrode systems.