• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권6호
• /
• pp.1545-1559
• /
• 2023

In the cloud environment, microservices are implemented through Kubernetes, and these services can be expanded or reduced through the autoscaling function under Kubernetes, depending on the service request or resource usage. However, the increase in the number of nodes or distributed microservices in Kubernetes and the unpredictable autoscaling function make it very difficult for system administrators to conduct operations. Artificial Intelligence for IT Operations (AIOps) supports resource management for cloud services through AI and has attracted attention as a solution to these problems. For example, after the AI model learns the metric or log data collected in the microservice units, failures can be inferred by predicting the resources in future data. However, it is difficult to construct data sets for generating learning models because many microservices used for autoscaling generate different metrics or logs in the same timestamp. In this study, we propose a cloud data refining module and structure that collects metric or log data in a microservice environment implemented by Kubernetes; and arranges it into computing resources corresponding to each service so that AI models can learn and analogize service-specific failures. We obtained Kubernetes-based AIOps learning data through this module, and after learning the built dataset through the AI model, we verified the prediction result through the differences between the obtained and actual data.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.559-561
• /
• 2022

인공지능 시대에 들어서면서 개인 맞춤형 환경을 제공하기 위하여 사람의 감정을 인식하고 교감하는 기술이 많이 발전되고 있다. 사람의 감정을 인식하는 방법으로는 얼굴, 음성, 신체 동작, 생체 신호 등이 있지만 이 중 가장 직관적이면서도 쉽게 접할 수 있는 것은 표정이다. 따라서, 본 논문에서는 정확도 높은 얼굴 감정 식별을 위해서 Convolution Block Attention Module(CBAM)의 각 Gate와 Residual Block, Skip Connection을 이용한 Identity- CBAM Module을 제안한다. CBAM의 각 Gate와 Residual Block을 이용하여 각각의 표정에 대한 핵심 특징 정보들을 강조하여 Context 한 모델로 변화시켜주는 효과를 가지게 하였으며 Skip-Connection을 이용하여 기울기 소실 및 폭발에 강인하게 해주는 모듈을 제안한다. AI-HUB의 한국인 감정 인식을 위한 복합 영상 데이터 세트를 이용하여 총 6개의 클래스로 구분하였으며, F1-Score, Accuracy 기준으로 Identity-CBAM 모듈을 적용하였을 때 Vanilla ResNet50, ResNet101 대비 F1-Score 0.4~2.7%, Accuracy 0.18~2.03%의 성능 향상을 달성하였다. 또한, Guided Backpropagation과 Guided GradCam을 통해 시각화하였을 때 중요 특징점들을 더 세밀하게 표현하는 것을 확인하였다. 결과적으로 이미지 내 표정 분류 Task에서 Vanilla ResNet50, ResNet101을 사용하는 것보다 Identity-CBAM Module을 함께 사용하는 것이 더 적합함을 입증하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권11호
• /
• pp.2987-3002
• /
• 2023

In 2015, the number of senior citizens aged 65 and over in Korea was 6,662,400, accounting for 13.1% of the total population. Along with these social phenomena, risk information related to the elderly is increasing every year. In particular, a fall accident caused by a fall can cause serious injury to an elderly person, so special attention is required. Therefore, in this paper, we implemented a system that monitors fall accidents and informs them in real time to minimize damage caused by falls. To this end, beacon-based indoor location positioning was performed and biometric information based on an integrated module was collected using various sensors. In other words, a multi-functional sensor integration module was designed based on Arduino to collect and monitor user's temperature, heart rate, and motion data in real time. Finally, through the analysis and prediction of measurement signals from the integrated module, damage from fall accidents can be reduced and rapid emergency treatment is possible. Through this, it is possible to reduce the damage caused by a fall accident, and rapid emergency treatment will be possible. In addition, it is expected to lead a new paradigm of safety systems through expansion and application to socially vulnerable groups.

• 로봇학회논문지
• /
• 제19권1호
• /
• pp.65-70
• /
• 2024

Tool wear is considered an important issue in manufacturing and engineering, as worn tools can negatively impact productivity and product quality. Given that the wear status of tools plays a decisive role in the production process, measuring tool wear is a key task. Consequently, there is significant attention in manufacturing fields on the precise measurement of tool wear. Current domestic methods for measuring wear are limited in terms of speed and efficiency, with traditional methods being time-consuming and reliant on subjective evaluation. To address these issues, we developed a measurement module implementing the DeepContour algorithm, which uses image processing technology for rapid measurement and evaluation of tool wear. This algorithm accurately extracts the tool's outline, assesses its condition, determines the degree of wear, and proves more efficient than existing, subjective, and time-consuming methods. The main objective of this paper is to design and apply in practice an algorithm and measurement module that can measure and evaluate tool wear using image processing technology. It focuses on determining the degree of wear by extracting the tool's outline, assessing its condition, and presenting the measured value to the operator.

• Current Photovoltaic Research
• /
• 제7권4호
• /
• pp.111-120
• /
• 2019

The output power of a crystalline silicon (c-Si) photovoltaic (PV) module is not directly the sum of the powers of its unit cells. There are several losses and gain mechanisms that reduce the total output power when solar cells are encapsulated into solar modules. Theses factors are getting high attention as the high cell efficiency achievement become more complex and expensive. More research works are involved to minimize the "cell-to-module" (CTM) loss. Our paper is aimed to focus on electrical losses due to interconnection and mismatch loss at PV modules. Research study shows that among all reasons of PV module failure 40.7% fails at interconnection. The mismatch loss in modern PV modules is very low (nearly 0.1%) but still lacks in the approach that determines all the contributing factors in mismatch loss. This review paper is related to study of interconnection loss technologies and key factors contributing to mismatch loss during module fabrication. Also, the improved interconnection technologies, understanding the approaches to mitigate the mismatch loss factors are precisely described here. This research study will give the approach of mitigating the loss and enable improvement in reliability of PV modules.

• 한국진공학회:학술대회논문집
• /
• 한국진공학회 2013년도 제44회 동계 정기학술대회 초록집
• /
• pp.498-498
• /
• 2013

Atmospheric Pressure Plasmas have pioneered a new field of plasma for biomedical application bridging plasma physics and biology. Biological and medical applications of plasmas have attracted considerable attention due to promising applications in medicine such as electro-surgery, dentistry, skin care and sterilization of heat-sensitive medical instruments [1]. Traditional approaches using electronic devices have limits in heating, high voltage shock, and high current shock for patients. It is a great demand for plasma medical industrial acceptance that the plasma generation device should be compact, inexpensive, and safe for patients. Microwave-excited micro-plasma has the highest feasibility compared with other types of plasma sources since it has the advantages of low power, low voltage, safety from high-voltage shock, electromagnetic compatibility, and long lifetime due to the low energy of striking ions [2]. Recent experiment [2] shows three-log reduction within 180-s treatment of S. mutans with a low-power palm-size microwave power module for biomedical application. Experiments using microwave plasma are discussed. This low-power palm-size microwave power module board includes a power amplifier (PA) chip, a phase locked loop (PLL) chip, and an impedance matching network. As it has been a success, more compact-size module is needed for the portability of microwave devices and for the various medical applications of microwave plasma source. For the plasma generator, a 1.35-GHz coaxial transmission line resonator (CTLR) [3] is used. The way of reducing the size and enhancing the performances of the module is examined.

• 상하수도학회지
• /
• 제31권4호
• /
• pp.303-310
• /
• 2017

Pressure retarded osmosis(PRO) has attracted much attention as potential technology to reduce the overall energy consumption for reverse osmosis(RO) desalination. The RO/PRO hybrid process is considered as the most logical next step for future desalination. The PRO process aims to harness the osmotic energy difference of two aqueous solutions separated by a semipermeable membrane. By using the concentrated water(RO brine) discharged from existing RO plants, the PRO process can effectively exploit a greater salinity gradient to reduce the energy cost of processing concentrated water. However, in order to use RO brine as the draw solution, PRO membrane must have high water flux and enough mechanical strength to withstand the high operational pressure. This study investigates the development of a thin film composite PRO membrane and spiral wound module for high power density. Also, the influence of membrane backing layer on the overall power density was studied using the characteristic factors of PRO membranes. Finally, the performance test of an 8-inch spiral wound module was carried out under various operating conditions(i.e. hydraulic pressure, flow rate, temperature). As the flow rate and temperature increased under the same hydraulic pressure, the PRO performance increased due to the growth of water permeability coefficient and osmotic pressure. For a high performance PRO system, in order to optimize the operating conditions, it is highly recommended that the flow pressure be minimized while the flow rate is maintained at a high level.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.563-570
• /
• 2022

Industrialized and modular construction is a growing construction technique that can transfer a large portion of the construction process to off-site fabrication yards. This method of construction often involves the fabrication, pre-assembly, and transportation of massive and long volumetric modules. The module weight keeps increasing as the modules become more complete (with infill) to minimize the work at the site and, as higher productivity can be achieved at the fabrication shop. Thus, a volumetric module delivery gets more challenging and risky. Despite its importance, past research paid relatively insufficient attention to the problem related to the lifting of heavy modules. This can be a complex and time-consuming problem with multiple lifting for transportation-and-installation operations both in fabrication yard and jobsite, and require complex crane operations (sometimes, more than one crane) due to crane load capacity and load balance/stability. This study investigates this problem by focusing on the structural perspective of lifting such long volumetric modules through simulation studies. Various scenarios of lifting a weighty module from the top using four lifting cables attached to crane hooks (either a single crane or double crane) are simulated in SAP software. The simulations account for various factors pertaining to structural indices, e.g., bending stress and deflection, to identify a proper method of module lifting from a structural point of view. The method can identify differences in structural indices allowing identification of structural efficiency and safety levels during lifting, which further allows the selection of the number of cranes and location of lifting points.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권4호
• /
• pp.1234-1257
• /
• 2023

In this paper, we propose a framework for multi-focus image fusion called PATN. In our approach, by aggregating deep features extracted based on the U-type Transformer mechanism and shallow features extracted using the PSA module, we make PATN feed both long-range image texture information and focus on local detail information of the image. Meanwhile, the edge-preserving information value of the fused image is enhanced using a dense residual block containing the Sobel gradient operator, and three loss functions are introduced to retain more source image texture information. PATN is compared with 17 more advanced MFIF methods on three datasets to verify the effectiveness and robustness of PATN.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.11-13
• /
• 2023

의미론적 분할(Semantic Segmentation)은 이미지 내의 객체 및 배경을 픽셀 단위로 분류하는 작업으로 정밀한 탐지가 요구되는 분야에서 활발히 연구되고 있다. 기존 어텐션 기법은 의미론적 분할의 다운샘플링(Downsampling) 과정에서 발생하는 정보손실을 완화하기 위해 널리 사용됐지만 고정된 Convolution 필터의 형태 때문에 객체의 형태에 따라 유동적으로 대응하지 못했다. 본 논문에서는 이를 보완하고자 Deformable Convolution과 셀프어텐션(Self-attention) 구조기반 어텐션 모듈을 사용한 의미론적 분할 모델을 제안한다.